Annals
of the
Missouri
Botanical
amen,
umber
Volume 89, Number 1
Winter 2002
Annals of the
Missouri Botanical Garden
The Annals, published quarterly, contains papers, primarily in systematic botany,
contributed from the Missouri Botanical Garden, St. Louis. Papers originating out-
side the Garden will also be accepted. All manuscripts are peer-reviewed by quali-
fied, independent reviewers. Authors should write the Managing Editor for informa-
tion concerning arrangements for publishing in the ANNALS. Instructions to Authors
are printed in the back of the last issue of each volume and are also available online
at www.mobot.org/mbgpress.
Editorial Committee
Victoria C. Hollowell
Editor,
Missouri Botanical Garden
Amy Scheuler McPherson
Managing Editor,
Missouri Botanical Garden
Diana Gunter
Associate Editor,
Missouri Botanical Garden
Aida Kadunic
Senior Secretary
Barbara Mack
Administrative Assistant
Ihsan A. Al-Shehbaz
Missouri Botanical Garden
Gerrit Davidse
Missouri Botanical Garden
Roy E. Gereau
Missouri Botanical Garden
Peter Goldblatt
Missouri Botanical Garden
Gordon McPherson
Missouri Botanical Garden
P. Mick Richardson
Missouri Botanical Garden
Henk van der Werff
Missouri Botanical Garden
For subscription information contact ANNALS
OF THE MISSOURI GARDEN, % Allen Marketing
. & Management, P.O. Box 1897, Lawrence, KS
.66044-8897. Subscription price for 2002 is
$140 per volume U.S., $150 Canada & Mexico,
$175 all other countries. Four issues per vol-
ume. The journal Novon is included i in the sub-
scription price of the ANNALS.
annals@mobot.org (editorial queries)
http://www.mobot.org/mbgpres 8.
© — Botanical Garden 2002
THE ANNALS OF THE MISSOURI BOTANICAL
GARDEN (ISSN 0026-6493) is published quar-
terly by the Missouri Botanical Garden, 2345
Tower Grove Avenue, St. Louis, MO 63110.
Periodicals postage paid at St. Louis, MO and
additional mailing offices. PosrMAsTER: Send
address changes to ANNALS OF THE MISSOURI
BOTANICAL GARDEN, % Allen Marketing &
Management, P.O. Box 1897, Lawrence, KS
66044-8897
The mission of the Missouri Botanical Garden is to disc "over and share — about plants and
thoi
their and enrich life,
LS
ANSI/ 150 239.48-1992 — — of Paper).
Volume 89 Annals
Number 1 of the
2002 Missouri
Botanical
Garden
A SYSTEMATIC REVISION OF Sylvain G. Razafimandimbison??
BREONIA (RUBIACEAE-
NAUCLEEAE)'
ABSTRACT
Breonia A. Rich., a Malagasy endemic genus of the tribe Naucleeae (Rubiaceae, Cinchonoideae), is revised here. The
morphology of Breonia species is described and compared with that of the other Malagasy Naucleeae, Breonadia Ridsdale.
Gyrostipula J.-F. Leroy, and Janotia J.-F. Leroy. Breonia is кизү gu from de 'se genera by having flattened and
elongated placentae, imbricate ovules, and indehiscent, multiple fruits r than triangular placentae, ovules attached
side by side to the base of the placentae, and free, lo fruits as * these genera. Neobreonia Ridsdale is again
included in Breonia. Twenty species are recognized, including eight that are new to science. A full taxonomic treatment,
keys, distribution maps of the recognized species are provided. The newly described species are illustrated.
Key words: | accrescent disk, Breonia, Cinc egit floral nectary. Mader! ar, Naucleeae, Rubiaceae.
Breonia A. Rich. is a Malagasy endemic genus swampy forests; they are absent from the semi-arid
of the tribe Naucleeae (Cinchonoideae, Rubiaceae) regions of southern Madagascar. The present revi-
and is the most diverse member of Naucleeae there. sion recognizes 20 species within Breonia, of which
Species of Breonia are large trees, or rarely shrubs, 8 are newly described herein. Breonia is distin-
that occupy habitats ranging from eastern rainfo- — guished from the other Malagasy Naucleeae, Breon-
rests to western deciduous dry forests as well as adia Ridsdale, Gyrostipula J.-F. Leroy, and Janotia
! | thank Charlotte Taylor for guiding this revision and generously sharing her knowledge of Rubiaceae with me; my
sicher Mick Ric — Pete Lowry, George Schatz, and James Miller for their advice and support кн my stay in
St. Louis; Peter Stevens, Toby Kellogg, three anonymous reviewers, and all the members of the Desmond Lee
molecular systematics lab of the Biology Department of the University of Missouri-Saint Louis A their invaluable
comments; Roy Gereau for his help with the Latin diagnoses; Petra De Block, Diane Bridson, and Aaron Davis for
their assistance and advice; Trisha Consiglio for her assistance with ArcView; Kendra Sikes for her assistance with the
TROPICOS database; David Frodin for sharing his thoughts on g Cephalanthus dient issues; Raphael Govaerts
for sharing his opinions on the Hichard and De Candolle issues; Barbara Alongi for her excellent illustrations; MEF
and ANGAP for collecting permits; and the following herbaria and pan staff for providing loans, access to collections,
and/or assistance in the field: BR, K, L, MO, P, PRE, TAN, ТЕК, and WAG. This research was supported by the
Ire
Andrew W. Mellon Foundation, the Liz Claiborne Foundation, the Rockefeller Foundation, the Garden Club of Alle ghe пу
County m and the Missouri Botanical Garden. This work was part of my Ph.D. dissertation presented to the Faculty
of the Gra rmm School of the University of Mis souri-Suin Louis, U.S.A.
? Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri 63166,
* Current address for reprint requests: fae of Systematic ем. — kis Centre, University of
Uppsala, Norbyv. 18D, SE-752 36, Uppsala, Sweden. sylvain. razafimandimbison@ebe.uu
ANN. Missouni Bor. GARD. 89: 1—37. 2002.
Annals of the
Missouri Botanical Garden
J.-F. 1
tae,
eroy, by having flattened, elongated placen-
imbricate ovules, and indehiscent, multiple
fruits rather than triangular placentae, ovules at-
tached side by side to the base of the placentae,
and free, capsular fruits as in these genera. Evi-
dence from recent molecular studies (Razafiman-
2002) showed that the mono-
typic genus Neobreonia Ridsdale and Breonia sensu
dimbison & Bremer,
Ridsdale together formed a monophyletic group. A
more recent phylogenetic study by Razafimandim-
bison and Bremer (in prep.) based on four data sets
(ITS, rbcL, trn T-L-F.
Neobreonia within Breonia, making the latter par-
and morphological) placed
aphyletic. Both genera have multiple fruits and
large accrescent disks. In light of this evidence, I
here merge Neobreonia with Breonia. Both studies
also showed that Breonia is more closely related to
Breonadia, Gyrostipula, and Janotia than it is to
the rest of Naucleeae.
Recognition of Breonia as a separate genus 15
1975).
its circumscription and species limits have been
1897; Homolle, 1938:
Ridsdale, 1975). Many more specimens were avail-
widely accepted (e.g.. Ridsdale, However,
controversial. (Haviland,
able for the present study than for earlier studies.
Ridsdale's
specimens of Breonia in the Paris herbarium (P).
revision was based primarily on. the
However, many of the specimens from the Forestry
herbarium in Madagascar (TEF) studied here are
not held as duplicates in P. Therefore, Ridsdale did
not see them.
MATERIALS AND METHODS
This revision is based on the examination of over
200 herbarium specimens and on field observations
of 9 of the 20 Breonia species. Specimens from the
following herbaria were examined and annotated
during this study: BR, K, L, MO, P, PRE, TAN, and
TEF (see Appendix 1). “S
All specimens with “SF”
meaning “Station Forestière,
апа “RN” meaning
“Réserves Naturelles,” were collected by Malagasy
technicians under the supervision of the French
I used the “SF” and “RN”
numbers rather than the collector numbers because
botanist Réné Capuron.
many specimens of this series do not have collector
names on their labels. Also, these are the numbers
used in the TROPICOS database of the Missouri
Botanical Garden.
Fieldwork was conducted in Madagascar in
May-July of 1995 and 1996, respectively, and Jan-
uary—June of 1998 to collect herbarium specimens
and pickled material of Breonia species, as well as
to gather ecogeographical data. The pickled mate-
rial was preserved in formalin/acetic acid/ethanol,
Table Distinctive fertile morphological features
separating Breonia and Sarcocephalus.
Characters Breonia Sarcocephalus
Inflorescence posi- axillary terminal
lon
Involucral bract always sur- never surround-
rounding the ing the young
young inflo- inflorescence
escence
Attachment of pla- to the septal to the middle of
cenlae to septa apices the septa
Stigma shape globose, clavate, spindle-shaped
capitate
ca. basal
Receptive areas on са. apical
the stigmatic
obe
Accrescent disks present absent
Number of ovules ] 10 9 more than 50
per locule
FAA (Radford et al., 1973). This allowed detailed
studies of inflorescences and infructescences and
revealed a number of important characters, which
are impossible to observe in dried specimens.
TAXONOMIC HISTORY
IDENTITY OF BREONIA
Breonia was originally described by Achille Ri-
chard (De Candolle, Sep. 1830); he named the ge-
nus after the early 19th century French botanist
Jean Nicolas Bréon, one of the first collectors. In
1879, Baillon placed Breonia in the synonymy of
Sarcocephalus Afz. ex Sabine solely because both
have multiple fruits, a character that evolved at
least three times in Naucleeae (Razafimandimbison
& Bremer, 2002). Baillon included two species of
Sarcocephalus, S. Rich.)
Baill. from Madagascar.
Breonia is quite distinct from Sarcocephalus in
many aspects (see Table 1). Schumann (1891),
Haviland (1897), and Homolle (1938) all endorsed
the recognition of Breonia as a separate genus. The
madagascariensis (A.
and S. richardiana Baill.,
separation of Breonia and Sarcocephalus and the
monophyly of Breonia is strongly supported by ITS
and rbcL sequence data (Razafimandimbison &
Bremer, 2002).
AUTHORSHIP OF BREONIA
The authorship of Breonia ч bee n cited previ-
Rich.
1975, 1978). I disagree with a pes Pyramus
de Candolle asked Achille Richard for a copy of his
manuscript of the Mémoire de la Famille Rubiacées
ously as Breonia A. . (see Ridsdale,
Volume 89, Number 1
2002
Razafimandimbison 3
Revision of Breonia
Table 2.
Species circumscriptions of Breonia.
(Richard, Dec. 1830) to avoid any possible nomen-
clatural conflicts between their work. At that time,
Richard’s work was already in press for publication.
Richard kindly sent De Candolle the manuscript,
along with 45 illustrations and later the page proofs,
hoping that his work would come out before De Can-
dolle's Prodromus Systematis Naturalis (Stearn,
1957). De Candolle was able to cite Richard’s work
ahead of its actual issue, since the Prodromus was
published in September 1830 ahead of Richard’s
book (published in December 1830). Comparison of
Richard’s original description of Breonia with that
included in De Candolle’s book showed that De Can-
dolle used Richard’s description of Breonia, but in
a different format. The authorship of the genus
should remain with Richard. Therefore, I cite Breon-
ia as Breonia A. Rich. in [DC.].
PREVIOUS SPECIES DELIMITATION IN BREONIA
The species limits used by previous authors have
been controversial. Haviland (1897), in his first
worldwide revision of the tribe Naucleeae, recog-
nized 10 species of Breonia (Table 2). He used pet-
iole length and secondary leaf vein number to di-
vide the genus into two major groups: the first group
Haviland (1897) Homolle (1938) Ridsdale (1975) Razafimandimbison (2001)
boivinii boivinii boivinii boivinii
coriacea —! not mentioned —*
cuspidata cuspidata — cuspidata
longipetiolata dubious species dubious species dubious species
madagascariensis madagascariensis madagascariensis madagascariensis
mauritiana —! — —
membranacea membranacea — membranacea
parvifolia parvifolia sphaerantha sphaerantha
richardiana richardiana — —
stipulata —! stipulata
decaryana Neobreonia decaryana decaryana
havilandiana — havilandiana
keliravina — —
louvelii — louvelii
macrocarpa — macrocarpa
Jerrleri perriert errieri
citrifolia chinensis
capuronil
fragifera
owryi
richardsonii
sambiranensis
—3 taolagnaroensis
tayloriana
Isaratananensis
—' merged in B. richardiana; — merged in В. madagascariensis; —? merged in B. citrifolia; — merged in B.
chinensis; —^ merged in B. decaryana.
comprising Breonia madagascariensis, with subses-
sile leaves and 12 secondary veins (there are ac-
tually 18 or 19 secondary veins, but the leaf blade
of the type specimen is folded, so Haviland must
have counted what he could see); and the second
group containing the remaining species, which had
petiolate leaves with 9 to 11 secondary veins. Hav-
iland then used leaf blade length and width, as well
as inflorescence width, to delimit the species of his
second “petiolate” group. Forty years later, Hom-
olle (1938) adopted new species circumscriptions
for Breonia. She retained six species from Havi-
land's treatment, but also described six new species
(see Table 2). Unfortunately, she did not provide a
key to the species she recognized. Ridsdale (1975)
proposed new circumscriptions for Breonia based
on a combination of the following characters: stip-
ules of terminal vegetative buds conical and ob-
volute; inflorescences always lateral; calyptra-like
bracts coherent, surrounding the young inflores-
cence, and circumscissile; adjacent calyx tubes free
and persistent; calyx lobes densely pubescent; and
placentae attached to the upper third of the septum.
He used the shape and length of stipules of ter-
minal vegetative buds, the mature stipule length,
Annals of the
Missouri Botanical Garden
and the leaf shape and size as primary characters
and recognized only five species (Table 2). He
placed Breonia decaryana Homolle and B. kelira-
vina Homolle in his new genus Neobreonia because
of their partly fused calyx tubes and complanate
terminal vegetative buds.
Capuron (1973a),
deposited at P, recognized 13 species using varia-
in an unpublished manuscript
tion in leaf blade size, stipule shape and length.
calyx tube length and indumentum, ovule number,
and fruit surface. However, he provided neither de-
scriptions of the new species that he recognized nor
a key for their identification.
ON THE IDENTITY OF CEPHALANTHUS CHINENSIS LAM.
AND ANTHOCEPHALUS А. RICH.
This section presents the problem of the identity
of Cephalanthus chinensis Lam. |= Breonia chinen-
sis (Lam.) Capuron] and its historical confusion with
Breonia. This issue has caused a long and contro-
versial debate during the last 30 years (cf. Bak-
huizen van den Brink, 1970; Capuron, 1973b;
Ridsdale, 1975; Bosser, 1984, 1999),
Lamarck (1785: 678) claimed that a specimen
he received from Sonnerat was the type of his Ce-
phalanthus chinensis. On the other hand, Richard
(Sep. 1830) also stated that he described his An-
thocephalus |= Neolamarckia Bosser| as an Asian,
and not Malagasy, monotypic genus (Anthocephalus
indicus) based on specimens labeled “C. chinensis
Lam.” Lamarck and Richard gave no further details
about their type specimens. Bakhuizen van den
Brink (1970) and Ridsdale (1975) surmized that the
collection of Sonnerat that served as the type of C.
chinensis was also used by Richard to describe his
Anthocephalus. Capuron (1973b) and Bosser (1984,
1999) argued that C.
phalus A. Rich. were not necessarily based on the
chinensis Lam. and Anthoce-
same specimen.
In the Lamarck herbarium in Paris (P-LA) there
is an authentic specimen of Cephalanthus with five
different labels attached to it. The first label is La-
о
marck’s and says ‚н chinensis enc n
The second label is in Commerson's handwriting
p
Bosser, 1984), with a dcn Latin description and
the country of origin mentioned as "Isle de France"
(i.e., Mauritius).
.. >
c'est la Morinda de Chine.” The fourth label
The third label is Sonnerat’s and
says
is also Sonnerat’s, on which he wrote “Morinda de
Chine," and below this name Lamarck added “Ce-
phalanthus.” The fifth label was written by Roeper
"Nauclea
orientalis Lam." This specimen sheet was consid-
ered by Capuron (1973b) and Bosser (1984, 1999)
and says "Nauclea purpurea Roxb.” and
to be the type of Lamarck’s Cephalanthus chinensis.
However, Bakhuizen van den Brink (1970) and
Ridsdale (1975) disagreed because this specimen
does not match Richard’s description of Anthoce-
phalus. Despite the fact that no one has ever found
LA that could
have served as the type of both C. chinensis Lam.
a collection of Sonnerat s.n. in P or P-
and Anthocephalus, Bakhuizen van den Brink and
Ridsdale inferred that Lamarck must have based
This collec-
his C. chinensis on such a collection.
tion, they further surmized, must contain two ele-
ments, one of Breonia and the other of Anthoce-
phalus. They also argued that Richard must have
typified his Anthocephalus indicus by only the ele-
ment of Anthocephalus of this hypothetical Sonnerat
collection. Bakhuizen van den Brink then typified
Anthocephalus by only the element of A. indicus of
this hypothetical Sonnerat collection, excluding its
element of Breonia. Ridsdale (1975) endorsed Bak-
huizen van den Brink's arguments and. chose the
unverified or non-existent collection as the lecto-
type of Anthocephalus.
Capuron (1973b), (1984,
999), disagreed with Bakhuizen van den Brink
(1970) and Ridsdale (1975). Capuron and Bosser ar-
gued that the type specimen of Cephalanthus chi-
followed by Bosser
nensis Lam. was the specimen with five labels in P-
LA because the original description of C. chinensis
am. (Lamarck, 1785: 678) — this plant, and
amarck himself annotated it. Bosser argued. that
amarck was apparently — ed by Sonnerat’s la-
bels stating “Morinda de Chine,” and assumed that
the plant originally grew in China, the Philippines,
1984: 244). Bosser then
inferred. that Sonnerat’s label meant that Sonnerat
and the Moluccas (Bosser,
had seen the specimen of C. chinensis now in P-LA,
and that it reminded him of what he called “Morinda
de Chine.” Bosser (1984) also argued that the sec-
ond label was Commerson's because there is a sim-
ilar specimen in P labeled C. chinensis Lam. and it
was collected by Commerson in Mauritius. In addi-
tion, the specimen with five labels in P-LA shows
axillary inflorescences and adjacent flowers with
fused ovaries that are characteristic of Breonia. Ca-
puron (1973b) then concluded that C. chinensis Lam.
was based on a plant originally from Madagascar
rather than Asia, and therefore belongs to Breonia.
Consequently, Capuron made the new combination
B. chinensis (Lam.) Capuron.
l agree with Capuron's and. Bosser's arguments.
A large number of living plant collections of early
French botanists (e.g., Chevalier Etienne de Fla-
Jean Nicolas Bréon,
court, Philibert Commerson,
André Michaux, Louis Armand Chapelier) were
sent to Mauritius for cultivation in the late 18th
Volume 89, Number 1
2002
Razafimandimbison 5
Revision of Breonia
century (Dorr, 1997). Commerson sent a large col-
lection of dried plant specimens to Paris, but these
specimens were labeled and distributed to various
1773 (Dorr,
1997). This may well have caused some confusion
herbaria only after his death on 13 May
between Commerson’s and Sonnerat’s specimens
[see. e.g.. Guého (1976) regarding the type of Phy-
Heine (1968) for Cordia sinensis
Lam.|; the type of C. chinensis is apparently another
lica nitida Lam.;
one of these cases.
Bosser (1984) has pointed out correctly that our
problem is to find out whether the specimen of C.
chinensis in P-LA conformed to what Lamarck said
in his original description of C. chinensis, rather
than what Richard said about Anthocephalus. In-
deed, Bosser argued that Richard may not have
seen the specimen of С. chinensis in P-LA because
Lamarck sold his herbarium to Roeper in 1824 and
the entire collection was not returned to Paris until
1886, while Richard’s work was in press for pub-
lication in 1830. Richard’s original description of
Anthocephalus indicates that he used specimens
with flowers and fruit, whereas Lamarck had
specimen with flowers only; therefore, even if Ri-
chard had seen the specimen in P-LA, he would
still have needed a specimen with fruit. Finally,
Anthocephalus, as Richard noted, has terminal in-
florescences, whereas the specimen of C. chinensis
in P-LA has axillary inflorescences. Therefore,
Bakhuizen van den Brink’s and Ridsdale’s scenar-
ios of C. chinensis being based on mixed elements
of Breonia of Sonnerat and Anthocephalus is unten-
able, in agreement with Bosser (1999).
Richard's protologue for Anthocephalus stated:
“Species observata: Anthocephalus indicus nob.
qpephatanthus chinensis Lam.—Nauclea purpurea
* (Richard, 1830: 157). Bosser argued that
the type specimen of B. mauritiana [= B. chinensis
herein] in P, collected by Commerson (s.n.) i
—
=)
Mauritius, may well have been one of the speci-
mens on which Richard based his Anthocephalus
because: (1) this specimen has two old labels; the
names on the second label are the same names as
those in the protologue of Anthocephalus; and (2)
there is a packet with pieces of an infructescence
attached to this specimen, and the separated indi-
vidual fruits of this infructescence match Richard’s
description of the infructescence of Anthocephalus.
Thus, Richard may well have used the fruiting head
attached to the type specimen of B. mauritiana as
part of his original description of Anthocephalus.
We know only that the element of Anthocephalus
was attached to the type sheet of B. mauritiana
when Haviland (1897) described this species: “
It is this specimen which has attached to it some
heads of Anthocephalus indicus and the curious
confusion about Nauclea purpurea and Cephalan-
thus chinensis.” | consider this fruiting head at-
tached to the type specimen of В. mauritiana as a
piece from the type specimen of Richard’s Antho-
cephalus indicus, which is now missing.
In conclusion, Cephalanthus chinensis Lam. is
Breonia, whereas Anthocephalus A. Rich. is Asian.
Bosser (1984) created a new generic name Neola-
marckia for the Asian taxa that match Richard’s
original description of Anthocephalus, and made the
new combination Neolamarckia cadamba (Roxb.)
Bosser based on Nauclea cadamba Roxb. (1824). |
agree with Bosser's combination because N. cadam-
ba Roxb. is the older name, and thus it has priority
over А. indicus A. Rich. Ridsdale
(1998), in his treatment of Rubiaceae for Sri Lanka,
continued to use Anthocephalus rather than Neo-
30). However,
lamarckia for these plants.
SPECIES CIRCUMSCRIPTIONS IN BREONIA SENSU
RAZAFIM.
Haviland's (1897) circumscription of Breonia spe-
cies based on secondary leaf veins, leaf blades, and
inflorescences is problematic because these charac-
ters overlap for some well-defined species. Using such
features as stipules of terminal vegetative bud size,
mature stipule length, and leaf shape and size as pri-
mary discriminating characters forced Ridsdale to
lump taxa and recognize only five species, of which
three (Breonia citrifolia, B. madagascariensis, and B.
sphaerantha) become rather heterogeneous. Neither
Haviland (1897) nor Ridsdale (1975) used floral and
fruit characters for their keys for Breonia. However,
there are a number of useful taxonomic characters,
such as shapes of the stipules of the terminal vege-
tative bud and inflorescence axis, corolla lobe indu-
mentum, the degree of fusion of adjacent ovaries, ac-
and infructescence surface a
| abandon Haviland's (1897) and Ridedale’s
1975) species delimitations and adopt here new spe-
crescent disks,
result,
—.
cies limits within Breonia.
My criterion for recognizing species is the pres-
ence of one or more apparently fixed or non-over-
lapping morphological differences between putative
species. Morphologically distinct groups of speci-
mens were identified and delimited by non-over-
lapping or fixed diagnostic differences; these di-
agnosable morphological units, which I refer to as
species, are hypotheses until new data are available
to refute them. To relate this functional species con-
cept to more theoretical ones, one must test it on
the basis of the criteria defining each theoretical
species concept.
Annals of the
Missouri Botanical Garden
=
С
б.
А B
р Е Е
Figure l. Variation in shape and size of the stipules of the vegetative buds in Breonia (A, С-В) апа Janotia.
Breonia decaryana.—B. Janotia macrostipula.—C.
'ariensis
MORPHOLOGY AND TAXONOMIC CHARACTERS OF
BREONIA
A detailed overview of the morphological char-
acters of Breonia species is presented as well as
their taxonomic significance. The allied Malagasy
genera Breonadia, Gyrostipula, and Janotia are also
discussed for comparison.
HABIT
Plants of Breonia are medium to large, some-
times emergent trees up to 30 m tall, rarely shrubs.
n Breonia sphaerantha.
Buttresses are reported
The bark is smooth to rugose. The trunk is typically
straight. The branches are usually plagiotropic.
LEAVES
Breonia leaves are simple, opposite, and decus-
sate, as in most Rubiaceae, and generally coria-
ceous to membranaceous. The length of petioles
and the size of leaf blades are useful for recognizing
some species. The leaves of Breonia are petiolate
or rarely subsessile (B. madagascariensis). Most
species have small to medium-sized leaves, but two
species (B. macrocarpa and В. madagascariensis)
the former to 38
The
have relatively large leaf blades:
X 25 em, and the latter up to 45 X 35 em.
leaf blades are mostly glabrous, or rarely pubescent
in a few individuals of B. perrieri. Breonia macro-
carpa is characterized by having brown, long hairs
on the lower surface of the leaf blades.
Domatia. | Domatia are cavities located in the ax-
ils of veins on the abaxial sides of leaf blades (Ja-
cobs, 1966). They typically occur in the axils of sec-
ondary veins in most species of Breonia. Mites or
mite eggs can be found in the domatia, although the
occurrence of domatia is not dependent on the
—A
B. capuronit.—D. B. chinensis. —E. B. stipulata.—V. B. madagas-
arachnid presence; these domatia may provide the
host plant protection from herbivores or pathogens
(Pemberton & Turner, 1989). The type and location
of domatia are useful for species delimitation i
Breonia. In B. macrocarpa, they occur in the axils
of the secondary and tertiary veins. Both secondary
and tertiary domatia have never been seen in other
Naucleeae; they are only known in a few members
of Rubiaceae (e.g.. Aoranthe, Pleiocoryne, Oligoco-
don) (Robbrecht, 1988). In Breonia, domatia can be
tufts (covered by dense hairs), or cryptic and. gla-
brous, or pits (depressions in the lamina with a broad
opening).
Venation. The venation of the leaf blades is
usually prominulous on the upper sides of the leaf
blades, but is prominent in Breonia havilandiana,
The
secondary venation is usually eucamptodromous
Radford et al., 197:
Stipules of — vegetative buds.
ules of terminal vegetative buds are usually conical
and obvolute in most Breonia species (Fig. 1C-F).
In these, two stipules overlap in the bud such that
В. madagascariensis, and B. tsaratananensis.
—
The stip-
one half of each is external and the other half is
internal (Harris € Harris, 1994).
lengths vary between species, which is useful for
their identification. In Breonia decaryana and Jan-
otia macrostipula, the stipules of terminal vegeta-
tive buds are complanate (Fig. 1A, B), the two stip-
ules being flat and pressed together. In Gyrostipula,
Their sizes and
they are long. filiform, and convolute, one stipule
being rolled inside another (Radford et al., 1973).
Mature stipules. The stipules of Breonia and
other Malagasy Naucleeae are interpetiolar, boat-
shaped (cymbiform), and usually abaxially carinate.
The keels are prominulous in many species and
prominent in B. tayloriana and В. madagascarien-
Volume 89, Number 1
Razafimandimbison 7
Revision of Breonia
In Gyrostipula, the stipules are without keels
and are twisted when dry. The size and persistence
of the stipules vary among species of Breonia and
are useful for species distinction. Stipules are typ-
ically deciduous in Breonia. In Breonia tayloriana,
they are semi-persistent.
INFLORESCENCES
Like other Naucleeae, Breonia species have glo-
bose inflorescences on which numerous small flow-
ers (usually more than 50) form dense clusters rath-
er than capitate other
Rubiaceae (e.g., Schradereae, Psychotrieae). The
individual flowers are inconspicuous, and the entire
cluster of flowers acts as a single attractive unit.
Inflorescence stalk. In Breonia, the inflores-
cence stalk is typically articulate and composed of
a peduncle, a node, and an inflorescence axis (e.g..
Figs. ӨК, OF, 14C). Ridsdale (1975) used the term
“flowering axes” to refer to what I call inflorescence
axes; he called “true peduncles” what I refer to as
inflorescences found in
peduncles. An evident node separates the peduncle
from the inflorescence axis. The inflorescence axis
is located between the branch bearing the inflores-
cence and the node. The peduncle is located be-
tween the node and the base of the head; this is
hidden by the bracts prior to the separation of the
bract lobes. The peduncle can be absent (e.g.. Figs.
15A, 16F); in some species (e.g., Fig. OF, Breonia
capuronii; B. decaryana), this continues to elongate
after the inflorescence axis has finished growing at
the end of anthesis. The length and shape of the
inflorescence axes are useful for recognizing spe-
cies in Breonia. Inflorescence axis shapes can be
terete (e.g... Вгеота саригопи, B.
chardsonii, B. sphaerantha) or flattened (e.g.. В.
chinensis, B. havilandiana, B. membranacea, B. sti-
pulata, and B. tayloriana). The inflorescence axes
are usually longer than the peduncles; however, in
Breonia decaryana, B. capuronii, and B. sphaer-
antha, the two are similar in size and shape.
racts. Most Breonia species have bracts that
are coherent, but histologically distinct (Radford et
al. 1973), valvate, and completely enclose the
young inflorescence in a calyptra-like fashion. As
the inflorescences develop, the bracts separate lon-
gitudinally into two lobes on the nodes lasting for
only a few days and usually falling off before an-
thesis. They may persist for two or three weeks after
anthesis in Breonadia, Gyrostipula, and Breonia de-
caryana. In B. richardsonii, the bracts are tubular,
appressed to the inflorescence axis, and terminated
by three to four broadly triangular lobes. They nev-
er enclose the young inflorescence. This is unique
in Breonia and the whole of Naucleeae. Finally, the
bracts in Breonia are usually glabrous, except in B.
macrocarpa and B. tsaratananensis, where they are
densely pubescent. The position of the bract varies
depending on whether or not the peduncle contin-
ues to elongate after the inflorescence axis ends its
elongation. When the peduncle does not elongate.
the bract lies immediately below the inflorescence,
as in B. chinensis, B. cuspidata, B. madagascarien-
sis, and В. taolagnaroensis.
Ridsdale’s (1975) descriptions of calyptra-like
bracts of Breonia being coherent, surrounding the
young inflorescence, and circumscissile, appear to
be incorrect. I have seen two herbarium specimens
of Breonia chinensis showing circumscissile calyp-
tra-like bracts. In several species, including B. chi-
В. havilandiana, В. macrocarpa, В.
branacea, and В. the
normally separate longitudinally into two equal
hemispheres. This apparent circumscissile ruptur-
ing of the bracts evidently happens during speci-
men preparation. The illustration of Sarcocephalus
richardii Drake [= Breonia chinensis (Lam.) Ca-
puron] in Grandidier (1897: 457) shows the circum-
scissile calyptra-like rupturing of the bract. It is
possible that Ridsdale (1975) was influenced by
this illustration.
Interfloral bracteoles are absent in Breonia, as
well as in Gyrostipula and Janotia. However, they
are present in Breonadia. In Breonia richardsonii
(Fig. 14F), the ovary bases are surrounded by long
hairs.
nensis, mem-
tsaratananensis, bracts
FLOWERS
The flowers in Breonia are actinomorphic, her-
maphroditic, protandrous, sessile, and usually 5-
merous, although 4-merous examples are occasion-
ally encountered. The flowers are typically 4-merous
in B. capuronii, B. decaryana, B. fragifera, and В.
sphaerantha.
Calyx. The calyx tubes (i.e., tubular parts of limb
above the hypanthium) in most Breonia species are
free from each other and clearly visible. In B. fragi-
fera and B. decaryana the calyces are barely evident.
These calyx tubes of adjacent flowers are only partly
fused in B. decaryana; they are completely fused in
B. fragifera. Therefore, the degree of calyx fusion is
useful for species recognition in Breonia.
The shape and surface of the calyx tubes are
useful at the species level. In Breonia chinensis, B.
lowryi, B. tayloriana, B. stipulata, and B. taolag-
naroensis, the calyx tubes are infundibular and
smooth. Other species have infundibular and
Annals of the
Missouri Botanical Garden
DENO 00
Ш ге 2.
. B. fragifera.—D. В. macrocarpa.—E. B. chinensis.—F.
ribbed calyx tubes. The ribs may be prominulous
(e.g, B. sambiranensis) or prominent (B. tsaratan-
anensis and B. boivinii). In B. tsaratananensis the
calyx tubes are constricted above the middle and
narrowed toward the base and the lobes; this is
unique in Breonia.
The shape of the calyx lobes varies from oblong
truncate apices.
to triangular, sometimes with
Breonia sambiranensis has a distinct calyx. lobe
shape: short, truncate, with a shallow depression in
the center and a short appendage on the edge to-
ward the style (Fig. 14D). Broadly triangular and
pubescent lobes characterize Gyrostipula, while
Breonadia and Janotia have long, linear, pubescent
obes.
Corolla. Тһе corolla of Breonia is always hy-
pocrateriform and vellow-tinged. In B. boivinii, B.
sambiranensis, and B. tsaratananensis, some of the
corolla lobes have dorsal protuberances (e.g.. see
B. sambiranensis, Vig. 14E). Corolla indumentum is
useful for species recognition. One species group
(B. chinensis, B. cuspidata, В. louvelii, B. mem-
branacea, B. stipulata, and B. taolagnaroensis) has
glabrous lobes. Breonia boivinii, B. sambiranensis,
and B. tsaratananensis all have pubescent or pu-
berulous lobes. In B. macrocarpa and B. madagas-
cariensis lobes are puberulous and marginally gla-
brate.
Nectaries. As in other members of Naucleeae
sensu Razafimandimbison and Bremer (2002), the
floral nectaries (i.e., disks) in Breonia are incon-
spicuous, surround the style base, and are embed-
ded in the hypanthia between the base of the co-
rolla tube and the top of the ovary. For all Breonia,
nectaries continue to grow during infructescence
development and become hardened and ме.
ous. | call such nectaries “accrescent disks.” Ca-
puron, in his unpublished treatment for Malagasy
Rubiaceae at P, also observed this type of disk in
Breonia but did not understand its nature. These
accrescent disks appear to be the single synapo-
morphy among members of Malagasy Naucleeae.
However, they tend to be much larger in Breonia
species than in Breonadia, Gyrostipula, and Jano-
Different shapes of accrescent disks in Breonia fruits. —A. Breonia decaryana.—B. B. sphaerantha.—
B. taolagnaroensis.
tia. The shape of these accrescent disks (Fig. 2)
varies from obconical to obtriangular, to pentagonal
or rounded. Their shape and size are useful for spe-
cies distinction. This is the first report of the use
of this character in the classification of Breonia as
well as in Rubiaceae.
Stamens. Breonia filaments are usually very
short and flattened; they are inserted in the throat
of the corolla tubes and sometimes slightly exsert-
ed. The anthers are always bicuspid at the base,
basifixed, introrse, dehiscing along longitudinal
slits.
Gynoecium. In Breonia, ovaries (hypanthia) of
individual flowers are always bicarpellate. They are
coherent and nonseptate in Breonia capuronii, В.
decaryana, B. fragifera, and B. sphaerantha (see
Figs. OD, ОР), whereas they are fused (syncarpous)
and septate in the remaining species. Adjacent ova-
ries are typically fused. However, in B. richardson-
п. prior to and during anthesis the ovaries are fused
only at the base, but as the development continues,
the fusion extends to the mid-point (post-genital fu-
sion), producing multiple fruits (Radford et al.,
1973). Ovule number varies from 1 to 9; all four
nonseptate and three of the septate species (B. cus-
B. louvelii, and B.
uniovulate; the. remaining 13 septate species are
pidata, taolagnaroensis) are
multiovulate.
P
Breonia.
acenta size and shape vary between species in
The uniovulate species have small and
ovate placentae (Fig. 9F), whereas the multiovulate
species have large, flattened, and elongated ones
(Fig. ЗС. D). The placentae are triangular in Breon-
adia, Gyrostipula, and Janotia (Fig. ЗА. B). Pla-
centae are usually apically attached to the septum,
and adnate to the lateral sides of the nectaries, lo-
cated in the distal center portions of the carpels
13H).
recognizing Breonia from the other Malagasy genera
(see Fig. These two characters are useful for
of Naucleeae, but have never been used for clas-
sification by earlier authors. Defining Breonia by
the placentae attached to the upper third of the
septum as originally done by Haviland (1897) in
Volume 89, Number 1
2002
Razafimandimbison 9
Revision of Breonia
M
Figure 3. Different shapes of placentae in гове
Breonadia salicina.—C. Breonia sambiranensis
0.1mm
his first worldwide revision of Naucleeae, and later
by Ridsdale (1975), appears to be incorrect.
Ovule attachment to the placenta is another use-
ful i that distinguishes Breonia from Breon-
adia, Gyrostipula, and Janotia. In Breonia, the
vules are flattened and imbricate (Fig. 4D) along
the length of the placentae; they are attached side
by side at the base of the placenta in Breonadia,
Gyrostipula, and Janotia (Fig. 4
Breonia stigmas are typically globose to clavate,
or rarely capitate, pollen presenters. The receptive
areas are always restricted to the top of the stig-
matic lobe.
INFRUCTESCENCES
In Breonia species the infructescences develop
from syncarpous ovaries to form multiple fruits that
are fleshy, almost woody, when dry. They are or-
namented by accrescent calyces in most species. In
B. fragifera and B. decaryana, the calyx remnants
are barely evident and their infructescence surfaces
are pusticulate (with small, broad, slight elevations
A
=
€
с
o B
ure 4 Ovule
8
salicina.
Ra
‚ D) and its allies (А, B). —A. Gyrostipula foveolata.—B.
—D. B. macrocarpa.
not high or abundant) and rugose, respectively. In
Breonia, all the fruits are fused and constitute mul-
tiple fruits; the individual fruits are berry-like,
bearing 1 to 9 seeds completely enclosed by thin
exocarps and thick, hard endocarps.
SEEDS
In Breonia, the seeds are usually strongly flat-
tened ellipsoid. Concavo-convex seeds are found in
Breonia chinensis and В. stipulata. Breonia seeds
are usually unwinged, although rudimentary wings
are sometimes observed at both ends (e.g., B. chi-
nensis and B. taolagnaroensis), or only at the base
(e.g., B. perrieri). The seeds are possibly released
when the corky accrescent disks fall off (see the
section on seed dispersal). These disks can be eas-
ily removed when the infructescences are still
fresh, or after soaking dried ones for several hours.
Seed-coat is lineate (e.g., Fig. 6C) in the nonseptate
species, and typically reticulate (e.g.. Figs. 9C,
11C, 14H, 16E) in the septate species.
4
NU
attac 'hments to d entae in — par its = (A-C). —A. Gyrostipula foveolata.—B. Breonadia
Annals of the
Missouri Botanical Garden
ECOLOGY
POLLINATION ECOLOGY
How Breonia pollinates remains unknown. In the
Ankarana Reserve (northern Madagascar) in 1998
| saw the beginning of anthesis in a young individ-
ual plant of B. perrieri. Flowers in each inflores-
cence opened simultaneously to emit ап intense
odor of nectar, which might attract visitors from
considerable distance. The plant was visited by
several honeybees for a few hours.
DIASPORE DISPERSAL
Entire infructescences. In Breonia, the entire in-
fructescence can function as a dispersal unit. In
1998, I observed in the Anjanaharibe Sud Reserve
(northeastern Madagascar) a number of mature in-
fructescences of B.
chinensis, a facultative rheo-
phyte, floating on the river. The corky accrescent
disks may act as flotation devices allowing the in-
fructescences to float until they reach the river-
bank, where they may be eaten by any frugivores.
After flower fall, the nectaries have a new function,
as cork-like seed releasers.
A different floating device has also been reported
on the seed-coat of the African Naucleeae Sarcoce-
phalus pobeguinit Pob. ex. Pell., which is a facultative
rheophyte (Abbiw, 1985). Abbiw collected mature
seeds of Nauclea diderrichii (De Wild. & T. Durand)
Merr. and Sarcocephalus latifolius (Smith) Bruce (both
terrestrial species
—
‚ as well as S. pobeguinii. A simple
experiment was performed by soaking the seeds of
each species in separate containers of water. After 24
hours, the water was poured off: seeds of Nauclea
diderrichii and S. latifolius lay at the bottom of the
containers, whereas those of S. pobeguinii were still
floating. Nauclea diderrichii and S. latifolius seeds
were noted to sink after only a few minutes of im-
mersion. The seed-coat apparently acts as a floating
device preventing the seeds from sinking.
Seeds. — Infructescences of Breonia are used
food by all species of Malagasy lemurs. In Lokobe
Reserve (Nosy Be District, Antsiranana Province, in
northern Madagascar), B. boivinii and B. sambira-
nensis were the second most important food item in
the diet of the black lemur (Lulemur macaco) in
early 1992. Thirteen percent and 25% of its feeding
time during January and February, respectively, were
spent on these two species of Breonia (Birkinshaw.
1995). Birkinshaw frequently observed intact seeds
of B. boivinii and B. sambiranensis in lemur drop-
pings. He found that 2 of 7 defecated seeds germi-
—
nated, compared with 7
=~
-
-30 seeds collected from
the ripe infructescences used as a control. The ma-
ture infructescences were not observed to be eaten
by any other frugivores in the Lokobe Reserve. Both
species of Breonia also occur in the Ambanja region
where Eulemur macaco is also known to be common,
suggesting that £. macaco may be an effective agent
of seed dispersal for B. boivinii and B. sambiranen-
s. Notably, B. sambiranensis is known from only two
specimens. outside of the Sambirano regions. Seed
dispersal mechanisms for other Breonia species re-
main unknown.
ECONOMIC USES
All Malagasy Naucleeae are known under the lo-
“Valotro,”
which is reserved for a particular use (Boiteau,
1985). From the 19th century through the mid 20th
century, the Malagasy Naucleeae were used for
cal name “Valotra” or meaning that
posts placed around villages to protect against en-
emies. All Breonia have various local uses such as
boats, bridges, making handerafts and furniture,
and for general construction purposes because they
produce high-quality, hard wood. Several species
are used medicinally. The infructescences of Breon-
ia are eaten by animals and humans.
SYSTEMATIC POSITION OF BREONIA SENSI
IN NAUCLEEAE SENSU RAZAFIM.
RAZAFIM.
& BREMER
Breonia has been placed in quite different tribes
over the last 170 years. Richard (1830) and De Can-
dolle (1830) placed Breonia in the tribe Opercular-
ieae along with other tribes with pluriovular locules.
Endlicher (1841) placed the genus in his subtribe
Sarcocephalinae (of the tribe Gardenieae), along with
Sarcocephalus, Zuccarinia Blume |= Jackiopsis
Ridsdale], Lucinaea DC. |= Schradera Vahl]. and
Canephora Juss. Schumann (1891), followed by Hav-
iland (1897), Verdcourt (1958), Ridsdale (1975), and
Robbrecht (1988, 1994), placed Breonia in the tribe
Naucleeae based on its compact spherical inflores-
Bremekamp (1966) transferred
Breonia, along with Adina Salisb., Mitragyna Korth.,
cences. However,
Uncaria Schreb., and Neonauclea Merrill, from the
tribe Naucleeae to the tribe Cinchoneae, arguing that
they differ from the other genera of Cinchoneae in
the capituliform inflorescence only. All these genera,
with the exception of Breonia, have capsular fruits
with winged seeds, which are characteristic of Cin-
choneae. Bremekamp provisionally placed these
genera in a separate subtribe within Cinchoneae
based on their capituliform inflorescences. Evidence
and rbcL
sequence data (Razafimandimbison & Bremer, 2002)
strongly suggests that Breonia belongs to Naucleeae
from recent molecular studies based on ITS
Volume 89, Number 1
2002
Razafimandimbison 11
Revision of Breonia
and is closely related to the other Malagasy genera
Breonadia, Gyrostipula, and Janotia. It is worth not-
ing that one species of African Uncaria, U. africana
var. africana, is also present in Madagascar. These
five genera can be identified using the key below.
Ккү TO THE GENERA OF MALAGASY NAUCLEEAE
la. Climbers, ре fang — — —
CUncadrid Se eh м species)
Ib Trees o or ere dnd fang pac abser
N
c
Hypanthia of the adjacent flowers fused or
at least at the base; fruits fused, indehiscent
Breonia A. Rich.
. Hypanthia of the adjacent flowers free; fruits
free, dehiscent.
3a. Leaves verticillate; stipules intrapetio-
lar: interfloral bracteoles present; seeds
unwingec
(20 pee ies)
3
- Breonadia Ridsdale (monotypic )
La eaves в бый stipules interpetiolar; in-
terfloral brac teoles 2 seeds winged.
4a. Leaf blades x 8-10 cm;
stipules of 8 vegetative
buds беде, — —
persistent — —
Janotia J. о. Leroy (monotypic)
Ab. Leaf biudes 8-10 X 3-5 cr i
ules of the terminal vegetative buds
convolute; stipules deciduous —
c
n; stip-
Gyosipila J- -F. Leroy (2 species, E ۳
ıich is yet to be described)
RELATIONSHIPS AMONG SPECIES
Species of Breonia can be subdivided into two dis-
(1) a
group with reduced calyces, coherent carpels, non-
tinct groups based on the following characters:
septate, and with lineate seed-coat (Breonia саригопи,
B. decaryana, B. fragifera, and B. sphaerantha); and
(2) a group with well-developed calyces, syncarpous
carpels, septate, and with reticulate seed-coat (re-
maining species). The monophyly of these two groups
needs to be tested using fast-evolving markers.
The included key to Breonia species does not re-
flect the above relationships; instead, | propose a
more practical key that puts more emphasis on the
most obvious unique vegetative character states, such
as the shape and arrangement of the stipules of the
terminal vegetative buds and size of leaf blades.
TAXONOMIC TREATMENT
Breonia A. Rich., in DC., Prodr. 4: 620. Sep. 1830.
T 4
РЕ: Breonia madagascariensis А. Rich
БЕ ТЕ" A. Rich.. Mém. Fam. Rubiacées. 210. Dec
YPE: Cephalidium citrifolium A. Rich. E
* onid —
Franchetia Baill., Bull. Mens Linn. Paris 1:
1885. YPE: Pains — Baill. [= B
sphaerantha |.
Elattospermum Sol., Bull. Herb. Boissier 1: 277. dad
TYP " lale Foe Sol. [= |
540. TYPE:
sphaerantha
майы "Ridsda le, Blumea 22: 1975.
Bre
reonia decaryana Homolle.
Trees or emergent trees, or rarely shrubs. Bark gray.
Stipules of terminal vegetative buds conical or rarely
ovate to obovate, obvolute or rarely complanate, gla-
brous or pubescent. Leaves simple, opposite, decus-
sate, persistent or deciduous; domatia present in axils
of secondary veins or rarely tertiary veins or absent:
stipules interpetiolar, mostly cymbiform or rarely com-
planate, deciduous or rarely semi-persistent, entire.
Inflorescences usually solitary or sometimes 2 or rare-
ly 4 to 8 per axil, axillary, globose; inflorescence axes
unbranched or rarely branched, flattened or terete,
usually slender or rarely robust and woody, glabrous
to pubescent, articulated or not; bracts usually calyp-
tra-like, cohering and completely enclosing the young
inflorescence, separating longitudinally into two equal
hemispherical shells, remaining attached to the node
for a few days and then falling off; peduncles elon-
gated (as internode above the inflorescence axis) or
not. Flowers hermaphroditic, radially symmetrical.
mostly 5-merous or sometimes 4-merous, closely con-
gested, sessile; calyx tubes infundibular, inside usu-
ally densely pubescent, outside glabrous except for a
few straight long hairs around the base, free from oth-
ers or rarely fused; calyx lobes oblong to truncate,
densely pubescent; corolla tubes hypocrateriform, in-
side glabrous to puberulous, outside glabrous; corolla
lobes oblong, glabrous or puberulous to pubescent,
aestivation imbricate in bud; stamens 5 or rarely 4,
inserted on the throat of the corolla tubes; anthers
introrse, partly exserted, dehiscing by longitudinal
slits, basifixed: filaments short, flattened, glabrous:
stigmas 1 per flower, clavate to globose or slightly
cylindrical, exserted from the corolla tubes, pollen
presenters; receptive areas restricted to the top of the
stigmatic lobe; ovary of an individual flower always
bicarpellate, syncarpous or coherent, adjacent ovaries
typically syncarpous or rarely fused at the base:
ovules | to 9 per locule, strongly flattened, pendulous,
imbricate; placentae apically attached to the septum,
usually elongated, flattened, adnate to the septum: flo-
ral nectary epigynous, embedded in the hypanthium
between the base of the corolla tube and the apex of
the ovary. Infructescences formed by multiple fruits,
globose, with persistent calyx remnants. Individual
fruits berry-like; endocarps hard and glossy or some-
times fibrous and soft; disks accrescent, conspicuous,
variable in size; seeds usually strongly flattened,
sometimes concavo-convex, or plano-convex, ellip-
soid, not alate or rarely with rudimentary wings, red:
seed-coat reticulate or lineate.
Annals of the
Missouri Botanical Garden
^
—
~
TO BREONIA SPECIES
Stipules of terminal vegetative buds flattened, complanate; inflorescences 4 to 8 per axil
›. В. decaryana
Stipules of terminal vegetative buds conical, кане infor 'scences usually ] or rarely 2 per г anil.
2a. Lower surfaces of leaf blades always pubescent, tuft-domatia present in the axils of secondary and
tertiary veins О. В. macrocarpa
2b. Lower surfaces of leaf blades always glabrous, tuft- domatia absent in the axils of secondary and tertiary
ve
за. Leaf blades 45 X 25 em; inflorescence axes 16-21 em long 1. B. madagascariensis
ЗЬ, e blades less than 32 X 16 i
› em; inflorescence axes less than 13 cm long.
la. Stipules semi-persistent 19. B. tayloriana
tb а deciduo
За af blades with the 3 secondary veins diverging from the base of the midrib, base
E late 3. perrieri
5b. Leaf blades without 3 secondary veins diverging from the base of the ШЕК, base
noncordate.
а. Bracts tubular, terminated by 3 or 4 broadly triangular lobes, never —
the young inflorescence, persistent; adjacent ovaries fused at the base
4. B. нш
Ob. Bracts e like. ешш; the young inflorescence, deciduous; adjacent
ovaries comple le ly fused.
Ta — ent c alyx "t completely fused: infructescences with barely evident
yx remnants ». В. fragifera
Adjac ent calyx tubes free: infructescences with well-developed calyx remnants.
a. C silla — pubescent to puberulous, recurved.
а.
af blades with — c i ic-Lype domatia on lower surface,
— n on upper surface; fertile peduncles densely pubescent;
calyx pubes unique ch dilated above the middle and constricted
at both ends 20. B. tsaratananensis
Leaf TM without domatia; fertile peduncles wlabrous: calyx
tubes funnel-shape
Ob.
10а. Lower surfaces of leaf blades yellow-red-tinged when dry:
calyx tubes 1-1.2 mm long; lobes 0.2-0.3 mm long, bearing
shallow depressions on the center and a short protuberance
n the edge toward the style 5. B. sambiranensis
10b. Lower surfaces brown-tinged when dry; calvx tubes 2.5-3
mm long: lobes 2-2.2 mm long, without alloy 6 'pres-
sions, without protuberance
8b. Corolla lobes glabrous and not recurve
Ila
. B. boivinii
nflorescence axes terete; ovaries m rent but not histologically
12a. Petioles I. 52 em long: secondary veins drying red-tingec
16. B. —
12b. — always more than 2 cm long; s sec ondary | veins
ng vellow-tinged
IIb. Inflore 'scence axes flattened; ovaries — st eie ie
13a. Stipules of terminal vegetative buds 15-22 mm one:
^ Leaf blades obovate, always wavy a п hy
. € —
В. havilandiana
14b. Leaf blades elliptic to lance latte, not wavy when dry
17. B. stipulata
13b. Stipules of terminal — tative buds 4-10 mm ling:
15a. Ovaries uniovule
loa. Leaf blades 9-16 X 3.5-6.5 em long, ovate to
oblong in shaj l . taolagnaroensis
lob. Leaf blades 4 0 X 1.5-3.5 em Tem oblanceo-
late in shape.
17a. Le af blade apices acute, эю cuneate;
fertile peduncles 2—4 cm lon
cc
3. B. louvelii
17b. Leaf blade apices санай; bases attenu-
ale; fertile — les 4.5-6.5 em long
. B. cuspidata
15b. Ovaries multiovulate 2 to 9 кита Pe r Р ule >).
18a. Petioles 3-6 mm long 2. B. membranacea
18b. Petioles at least 9 mm long.
Volume 89, Number 1
Razafimandimbison
Revision of Breonia
1. Hasen boivinii Havil., J. Linn. Soc. Bot. 33:
1897. TYPE: Madagascar. [Antsiranana
province], District Nosy be [without exact lo-
cality], Boivin s.n. (holotype, K!; isotype, P not
seen).
Shrubs, ca. 7 m tall. Bark rugose. Leafy stems
terete, Stipules of terminal vegetative
buds conical, 5— ca. 5 mm, glabrous. Leaves
persistent; petioles (10—)15—25 X ca. 4 mm, terete,
glabrous; blades 11.5-23 X 4.5-13.5 cm, ovate to
broadly obovate, glabrous, coriaceous, glossy. apex
base cuneate, lower surfaces
glabrous.
acute to rounded,
brown-tinged when dry; margins glabrous, entire:
secondary veins ca. 9 pairs per side, eucampto-
dromous: tuft-type domatia in the axils of midribs
or absent; stipules 6-9 X ca. 6 mm, cymbiform,
19a. Stipules terminal vegetative buds more
u m long; inside of calyx tubes
lox ule -
‚ B. chinensis
п.
ай 2—4 ovules per
19b. Stipules of ый уе — bude З тт
long; inside of calyx tubes glabrous; тоге
numerous ovules (7—9) per loc Ж
. B. lowryi
sometimes emarginate, not carinate, glabrous, free
at the base, deciduous. Inflorescences solitary;
heads 3.5-3.8 cm wide, including stigmas; inflo-
rescence axes 2—4(—6.7) cm long, quadrangular to
slightly flattened, glabrous; bracts calyptra-like, de-
ciduous; peduncles (2.5-5)15-25 mm long. Flowers
5-merous; calyx tubes 2.5—3 mm long, prominently
ribbed, glabrous, dilated above the middle, con-
stricted at both ends, lobes 2—2.2 mm long, trun-
cate, prominently ribbed, inside puberulous to to-
mentose, outside glabrous, toward the base of lobes
tomentose; corolla tubes 6-7 X ca. 0.5 mm, gla-
brous; lobes 3—3.1 mm long, oblong, recurved, cil-
iate, inside puberulous with a few scattered long
hairs around the apex, outside glabrous; anthers ca.
1.5 mm long; filaments ca. 0.5 mm long, glabrous,
50°
N
3c E
А boivinii :
14? ` 14°
© havilandiana
* macrocarpa
Ш рете
18° 18°
272° 229
0 200 Kilometers
50°
46°
Figure 5.
5. Distribution of Breonia boivinii, B. havilandiana, В. macrocarpa, and B. perrieri.
Annals of the
Missouri Botanical Garden
V
b
^
“2
S
CP
pA
UA Tus O
Өе
— M
N Б
SINS:
D
e
d
X
(°С
Figure 6. Breonia capuronit.—A. Fertile branch with infructescences. В. Fertile branch with inflorescences. —
C. Seed: dorsal view (left); lateral view (right). —D. Two adjacent flowers, showing two adjacent ovaries separated:
dissection through corolla and calyx tubes and an ovary of one flower (left); separate flower (right). —E. Median
Volume 89, Number 1
2002
Razafimandimbison 15
Revision of Breonia
*
14° А decaryana
yk fragifera
capuronii
200 Kilometers
46°
Figure 7
flattened: styles 8-9 mm long, stigmas capitate:
ovary 2-carpellate; carpels syncarpous; ovules 6
per locule, pendulous, imbricate; placentae flat-
tened. Infructescences 3-3.5 cm diam., with well-
developed calyx remnants; fruits with endocarp soft
and fibrous: disks accrescent, obconical; seeds 4 to
6 per locule, strongly flattened, ellipsoid, white-
tinged; seed-coat reticulate.
Habitat and distribution. Low- and mid-alti-
tude evergreen rainforests; Districts of Nosy be,
Ambanja (Sambirano regions) and Vohémar (Fig. 5).
Common name. — Valotro (which is reserved for
Flowering November to December;
fruiting January to February.
Additional specimens — MADAGASCAR.
Antsiranana: District Nosy be, Réserve Integrale de Lo-
kobe, Antilahimena 19 (MO). үзе шин. 44 (МО
2751 RN (Р), Birkinshaw 67 (К, МО, Р, TAN); АК.
T
Distribution of Breonia capuronii, B. decaryana, and B. fragifera.
1432 SF (TEF); Doanilahy, 4905 RN (TAN), 6234 RN
(ТЕЕ); District Ambanja, Baie d'Ambato, 23424 SF
(TEF). Randrianaivo 247 (MO), a s.n. (P): District
Vohémar, Canton Antsirabe- — Forêt d'Analamateza,
27593 SF (TEF), 27636 SF (TE
2. Breonia capuronii Razafim., sp. TYPE:
Madagascar. | Antsiranana province], Massif de
nov.
Montagne d'Ambre, around “Station forestière
des Roussettes,” ca. 800-1000 m, 15 Feb.
(fl, young infr), 22059 SF (holotype.
TEF). Figures 1C and 6
Haec species ad congeneros nonnullos loculis uniovu-
latis accedit, sed ab eis petiolo longo (semper 2 cm ex-
cedente) — foliorum nervis secundariis in sicco flavi-
dis distinguitt
Trees, 15-30 m tall. Bark white-tinged. rugose.
Leafy stems terete, puberulous to pubescent, len-
ticellate. Stipules of terminal vegetative buds con-
ical, 4-5 X 1.5-2 mm, puberulous. Leaves decid-
E
dissection of developed post-anthesal mature flower showing calyx remnant, accrescent disk, and the single seed per
locule (left); entire fruit (right). —F. Portion of branch apex with a mature infruc tescence, a stipule of terminal vegetative
bud, and two petioles. A-D from 22059 SF (TE F) and E, F from 78518 SF (TEF).
Annals of the
Missouri Botanical Garden
uous; petioles 22-35 X ca. 1 mm, terete, glabrous,
lenticellate; blades 9.2—11.5 X 5.3-7.8 em, broad-
ly elliptic to obovate, clustered near the stem api-
ces, glabrous, membranaceous, glossy, apex acute
to mucronate, base rounded; margins glabrous, en-
tire; secondary veins ca. 8 pairs per side, eucamp-
todromous, adaxially conspicuous, yellow-tinged;
domatia absent; stipules ca. 4 mm long, cymbiform,
not carinate, glabrous, free at the base, deciduous.
Inflorescence solitary; heads ca. 2 em wide; inflo-
rescence axes ca. 1.7 em long, twisted when dry,
terete, densely puberulous; bracts calyptra-like, de-
ciduous; peduncles ca. 1 mm long, densely pubes-
cent. Flowers typically 4-merous; calyx tubes ca. 1
mm long, green-yellow-tinged, inside velutinous,
outside pubescent, lobes 0.7-0.9 mm long, trian-
0.2-0.5
mm, glabrous; lobes ca. 1.5 mm long, oblong, gla-
gular, pubescent; corolla tubes 2.8-3.0 X
brous, toward the apex gradually puberulous, cili-
ate; anthers ca. 0.5-1 mm long; filaments ca. 0.2
7.8-8 X ca. 0.1
mm, glabrous; stigmas globose or clavate; ovary 2-
mm long, glabrous, terete; styles
carpellate; carpels coherent; ovule 1 per locule,
pendulous; placentae small, elongated. Infructes-
cences 6-10 mm diam., with persistent calyx rem-
nants; individual fruits with endocarp hard, glossy:
disks accrescent, obconical;
seed ] per locule,
strongly flattened, ellipsoid, red; seed-coat lineate.
Habitat and distribution. Mid-altitude ever-
green rainforests and semi-deciduous forests: Dis-
tricts of Antsiranana Il, Ambanja, Sambava, and
Tsiroanimandidy (Fig. 7)
Common names. Valodrano (Valotra growing in
water), Valotro.
Phenology. Flowering January-February: fruit-
ing March-April.
Discussion. This species is different from the
other species of Breonia with a single ovule per
locule by having long petioles (always more than 2
cm), and secondary leaf veins drying yellow-tinged.
It has a large geographical distribution, but appears
to be locally rare or perhaps under-collected due
to its large size. The type species was collected
within the Montagne d'Ambre National Park
boundaries.
The specific epithet honors Réné Capuron, a
French botanist based Madagascar for several
years in the mid 20th century who worked mainly
on Malagasy woody plant families (including Ru-
biaceae).
Paratypes. MADAGASCAR. Antananarivo: District
Tsiroanimandidy, Forêt d'Ambohijanahary, 18518 SF (Р,
TEF ntsiranana: Montagne d’Ambre National Park,
Malcomber et al. 1219 (MO); District Ambanja, Canton Ma-
>
Analamateza, 7906 RN (TEF), 12975 RN (TEF);
Manongarivo, Gautier 3772 (G, MO, ТАМ); District Sam-
ava, Canton Maroambihy, Andranomadiohely, 9002 RN
(TEF).
rovato,
3. Breonia chinensis (Lam.) Capuron, Adanso-
nia, sér. 2, 13: 473. 1973. Cephalanthus chi-
Méth. 1: 678. 1785.
). [Without
Commerson s.n. (holotype, P-
nensis Lam..
TYPE:
exact locality |,
Encycl.
"Isle de France" (Mauritius).
LA photo!). Figures ID and 2E.
Méth. 4: 435.
. Rich., Mém.
шшс citrifolia Poir., in Lam., Encycl.
789. C Н € (Poir.) A
* Rub ‚ Dec. Згеота citrifolia (Poir.)
Ridsdale, pe 22: Er 1975. TYPE: Madagas-
car. [Without exact um collector unknown s.n.
(holotype, Р-ГА phot
и Бенни Baill. Adansonia 12:
187 о" ric hardiana i l.) Havi Linn.
Soc. ey . 1897. TYPE: Madagascar. r [W ithout
ct loc — e der s.n. (holotype, P!)
— coriacea. Havil., J. Linn. Soc. Bot. 33: A 1897.
TY e E: an ipt ar. M ithout exact loo ality], Hum-
i. (holotype, К!)
Phat. mauritiana Mod. J. Linn. Soc. Bot. 33: 35.
1897. TYPE: Mauritius. [Without exact. loc АШУ].
Cümmarm s.n. (holotype, P!).
Sarcocephalus richardii Drake, in A. Grandidier, Hist. Pl.
denen ar 36: 457. 1897. prose richardii Pa-
lacky, Cat. P nd ar 4: 50. YPE: Gran-
didier А. 457 (not sec
—
Shrubs or trees, 7-25 m tall. Bark rugose. Leafy
stems quadrangular, glabrous. Stipules of terminal
5—0
brous. Leaves persistent; petioles 9-15 X ca.
vegetative buds conical, 1-1.3 mm, gla-
mm, adaxially canaliculate, glabrous; blades 6.4—
2.4-5.5 cm, obovate, glabrous, coriaceous,
not glossy, apex broadly to narrowly cuspidate, base
cuneate; margins glabrous, entire; secondary veins
8 to 11 pairs per side, eucamptodromous, slightly
prominulous; domatia absent; stipules 7-11 X 1.5-
2.5 mm, cymbiform, carinate, glabrous, free at the
base, deciduous. Inflorescence solitary; heads 1—
2.5 cm wide; inflorescence axes 3.4—6.5 cm long,
glabrous; bracts calyptra-like, deciduous; pedun-
cles 1-2 mm long. Flowers 5-merous; calyx tubes
ca. 1.5 mm long, inside lanate, outside glabrous,
pubescent toward the base of the lobes, lobes ca.
0.5 mm long, truncate, lanate; corolla tubes 2.5—5
X 0.5 mm, glabrous; lobes 1.5-2 mm long, oblong,
8
mm long, glabrous, flattened; styles 7.5—9
abrous; anthers 1—1.5 mm long; filaments ca. 0.1
х 0.3-
0.4 mm, glabrous; stigmas clavate to capitate; ovary
2-carpellate; carpels syncarpous; ovules 2 to 3 per
locule, pendulous, imbricate; placentae elongated.
Infructescences 0.7-2 cm diam., with well-devel-
oped persistent calyx remnants; individual fruits
with endocarp hard, glossy; disks accrescent, ob-
Volume 89, Number 1
2002
Razafimandimbison 17
Revision of Breonia
50°
N
14°] 5 14°
Ж chinensis
18° 18°
22° 22°
200 Kilometers
46° 50°
Figure 8. Distribution of Breonia chinensis.
conical; seeds 2 to 3 per locule, with rudimentary
wings at both ends, strongly flattened, concavo-con-
vex, ellipsoid, red; seed-coat reticulate.
Habitat and distribution. Evergreen rainforests
and mid-altitude humid forests, occasionally in dry-
land sites and riverbeds; Districts of Antalaha,
Sambava, Farafangana, Fort-Carnot, Ifanadiana,
Manakara, Mananjary, Vangaindrano, Vohipeno,
Brickaville, Mahanoro, Maroantsetra, Soanierana
Ivongo, and Vavatenina (Fig.
Common names. Molompangady (lips of Valo-
tra), Valopangady (spade of Valotra), Valotra, Valo-
potsy (white Valotra), Vavalotra, Voakiringy, and
Voamalopangady.
Phenology. Flowering August to October; fruit-
ing November to March
Discussion. Breonia chinensis is the most com-
mon species of Breonia in Madagascar but grows
only in evergreen rainforests. The epithet “chinen-
sis” is not the most appropriate choice for Breonia,
which is restricted to Madagascar; however, it has
priority over the other available basionyms. Can-
didate basionyms of B. chinensis were all consid-
ered separate species of Breonia until Capuron
(1973a), endorsed by Bosser (1984, 1999), clarified
the identity of Cephalanthus chinensis Lam. as a
separate species of Breonia. Homolle (1938) re-
duced Breonia coriacea Havil. to synonymy under
B. richardiana Havil. Breonia coriacea was not
mentioned by Ridsdale (1975), but both species def-
initely belong to Breonia chinensis based on their
leaf blade shape and size, inflorescence axis shape,
and ovule number.
It has been accepted that the original VAN d
of Nauclea citrifolia Poir. [= B. chinensis] is in
marck's Encyclopédie (1789: 435) and that this spe-
cies belongs to Breonia (Bakhuizen van den Brink,
1970; Ridsdale, 1975; Bosser, 1984, 1999). How-
ever, Poiret did not specify type specimens of any
of his species. This has caused some confusion as
to which specimen he based N. citrifolia on and led
to disagreement over the identity of the name.
Richard (1830: 219) stated that his Cephalidium
citrifolium (Poir.) A. Rich. was a new combination
based on N. citrifolia Poir.: “Nauclea sp. Poir., En-
cycl. méth., Species unica observata: Cephal-
idium citrifolium, Nob. (Nauclea citrifolia, Poiret,
. C.).” Like Poiret, Richard also did not cite the
type specimens of any of his species.
There is one specimen (lacking the collector's
Annals o
Missouri Botanical Garden
name and number) with four labels in P-LA that may
be the type of N. citrifolia Poir. The first label i
Lamarck’s and says: “Nauclea citrifolia dic n° 2^
The second is Roepers and states: “? 8. Nauclea
cadamba Roxb.? [and] Nauclea citrifolia Poir.!!!, per-
quem affinis orientalis! immonimis affinis." The third
states that this specimen returned to P-LA: Herbier
de Lamarck, Novembre 1886." The
fourth label is that of Capuron and states: “Breonia
chinensis (Lamk.) R. Capuron (1973a)
considered that this Lamarck specimen is the type
of Nauclea citrifolia Poir. and is conspecific with
Breonia chinensis (Lam.) Capuron. I concur with
him because the protologue of N. citrifolia Poir. in
Lamarck (1789:
Ridsdale (1975) also endorsed Capuron's view, but
made the new combination Breonia citrifolia (Poir.)
Ridsdale because he did not recognize Cephalan-
thus chinensis Lam. as a Breonia. Noteworthy is that
Ridsdale did not cite the type specimen of N.
trifolia Poir. Bosser (1984) disagreed with Capuron.
arguing that an unnumbered Commerson collection
from Madagascar in P-LA is the type specimen of
N. citrifolia Poir. It is unclear whether Bosser re-
ferred to the Commerson s.n. in P-LA that serves
as the type of Cephalanthus chinensis Lam. In any
case, this does not correspond to the protologue of
N. citrifolia Poir.
Despite Richards (Richard, 1830:
210) that his Cephalidium citrifolium |= Breonia
chinensis| was based on Nauclea citrifolia Poir.,
Bosser (1984) argued that the name Cephalidium
was not based on the same type as N. citrifolia and
therefore should be cited as Cephalidium citrifol-
ium A. Rich. Also,
type specimen of C. citrifolium is another unnum-
“acquis en
Capuron.”
135) agrees with this specimen.
statement
Bosser further argued that the
bered collection of Commerson now in P. This Com-
merson s.n. at P collection is misidentified as N.
citrifolia Poir., but it is actually a completely dif-
ferent species of Breonia | describe here as B. tao-
lagnaroensis. This P collection of Commerson has
fruit only, whereas both Poiret's protologue of N.
citrifolia and Richard’s protologue of Cephalidium
citrifolium included flowers and fruits. Ridsdale
(1975) treated Cephalidium citrifolium (Poir.) А.
Richard under his Breonia citrifolia (Poir.) Rids-
dale, but I here consider it as a synonym of Breonia
—
chinensis (Lam.) Capuron.
MADAGASCAR
“suivant le
Ambo-
е е
alaha, SEES
SF (ТЇ
Additional m
Antsiranana: District
hitralanana, 9982 Ri iud
TEF); ме ине Forêt de Bero, 12447 S jd (TEF); Am-
Vinanivao, Forêt de Sahaen-
N); Canton a Vi-
D
bodiazovola, pipes.
jika, Bernard 327 (MO, TA
b 238 SF
— e Pare Masoala, Bernard 187 (MO,
TAN); District Sambava, Canton Analamanara au Sud de
Antsirabe-Nord,
Vigreuse 15418
de 7 en Sambava and
; [Unknown locality |.
Tartana,
27192 SF (TE
ТЫ
( Fianarantsoa: District Farafangana, Karianga,
Decary 5511 (PRE), Decary 5513 т) Dec ary 5573 (PRE):
Efatsy, Forêt d? alia 15381 SF (TEF); Canton vain,
SF (TEF), Aedes ^ Dis-
Forêt d’ жашашып,
M
Forét d'Analila, 162
trict Fort-Carnot, ded
15287 SF (TEF); Canton Tolongoina,
5964 SF (TEF); Ambatomalama, 15518 (TEF); Canton
Manampatrana, "exploitation forestiere" yit lonila-
d FEE); District Ifanadiana, Ranomafana м
p ‘tween Morafeno and Sahavanana, Turk
" 643 ( TAN); Canton. Kianjavato, entre nalis
Ned жу, 23917 SF (TEF); District Manakara, Fort
E Andafa. 14728 SF (TEF); Forét de Manakara au d
du terrain aviation, 86-R-118 (TEF); District Mus no,
Ifanirea,
vestiges de forét po 1 Nord de Vohipe no, 23691 (TEF);
Tsararano, 6300 (TEF); District Vanno. Са ton
Anosimparihy, ab ana, 14437 SF (TEF); Canton Mor-
aleno, CES да. 16174 SF (TEF); Marofototra, For-
êt de Mananjavara, 14712 SF (TEF); District Vaingain-
drano, between Loy and агре 23661 SF
TEF). Toamasina: District Brickaville, Managisy, 12352
SF (TEF); I
district Mahanoro, Canton d' Ambodinanidilana
| km à l'Est d'Ambodiala, Forêt d'Amboagibe, 19658 SF
ТЕЕ); remnant native vege — 1-2 km E of eid
ambo, Schatz et al. 3858 ‚ TAN); District Maroantse-
tra, Canton. Andranofoy, Anata au bord de la ri-
vière d'Andranofotsy, 21596 SF (TEF); Baie d'Antongil,
bassin de — between TT and Beanana,
9047 SF National Park, Andrombazaha,
Raha rion el )
2410 SF (ТЕЕ);
—
os Il, Cant
asa Mahatsara. 33240 SF (TE F), < А 711 SF (TEF),
` (TEF); Ambodiriana, 17299 SF (TEF): Vohi-
marangitra, 3/809 (TE Ja Forêt d eet à l'Ouest de
Foulpointe, 28087 SF (ТЕК); District Vavatenina, Canton
Sahatavy, Maizinandro, 26703 SF (TEF).
4. Breonia cuspidata (Baker) Haviland, J. Linn.
Soc. Bot. 33: 37. 1897. Nauclea cuspidata Bak-
er, J. Linn. Soc. Bot. 25: 319. 1890. TYPE:
Madagascar. "North West" [without exact local-
ity]. Sep. 1887 (fl); Baron 5563 (holotype, K!).
Trees. height unknown. Bark with annular fis-
sures. Leafy stems rounded, glabrous. ‘Terminal
vegetative buds conical, 4-5 X 1-1.2 mm, gla-
10-13 mm
long. adaxially canaliculate, glabrous; blades 5.5-
10 х 1.9-3 em, oblanceolate, glabrous, membra-
naceous, glossy, apex caudate, base attenuate; mar-
brous. Leaves persistent; petioles c:
gins glabrous, entire; secondary veins 8 to 9 pairs
per side, eucamptodromous, adaxially inconspicu-
ous, abaxially conspicuous; domatia absent; stip-
ules ca. 5 mm long, cymbiform, not carinate, gla-
deciduous. Inflorescence
brous, free at the base,
solitary, heads ca. 2 cm wide: inflorescence axes
4.5-6.5 cm long, strongly flattened, slender, gla-
brous: bracts calyptra-like, deciduous: peduncles
Flowers 5-merous:
not present. calyx tubes ca. 1
Volume 89, Number 1
20
Razafimandimbison 19
Revision of Breonia
mm long, free, inside lanate, outside up to the mid-
dle glabrous, above the middle lanate; lobes ca. 0.2
mm long, triangular, lanate; corolla tubes ca. 5 X
0 lobes 1.1-1.3
long, glabrous; anthers 0.7-0.8 mm long, filaments
mm; mm long, broadly ob-
0.2-0.3 mm long; styles 8—9 mm long; stigmas cla-
vate to capitate; ovary 2-carpellate; carpels syncar-
pous; ovule 1 per locule, pendulous; placentae re-
duced, pendulous. Infructescences not seen.
Distribution and habitat. Northwest Madagas-
ar; habitats unknown
Common names. Unknown
Flowering September; fruiting time
Discussion. Breonia cuspidata appears to have
a restricted distribution and has not been collected
since 1890. This species is distinguished from the
other species by its ovaries with one ovule per loc-
ule and its leaf blades with caudate apices and at-
tenuate bases.
Original paratypes. e locality]. Baron 6602
(K!) and Scott Elliott 2214 (K!).
5. Breonia decaryana Homolle, Bull. Soc. Bot.
France 84: 460. 1937 [publ. 1938]. Neobreon-
ia dec aryana (Homolle) Ridsdale, Blumea 22:
540. 1975
province], ріне Farafangana, Ifandana, De-
:: Madagascar. [Fianarantsoa
cary 5199 (holotype, P!; isotype, L not seen).
Figures 1А and 2A.
Breonia keliravina Homolle, Bull. Soc. Bot. — 84:
460. 1937 [publ. 1938]. TYPE: Madagascar. Ana-
amazaotra, Thouvenot 91 (lectotype, es by
P!).
Ridsdale (1975: 546),
Trees, 10-30 m tall. Bark rugose with annular
fissures, rarely smooth, lenticellate. Leafy stems
quadrangular, glabrous, always dichotomously
branched. Terminal vegetative buds complanate, 8—
14 X 6-10 mm, glabrous. Leaves persistent; peti-
oles 10-20 X 2—3 mm, adaxially canaliculate, gla-
leaf blades (4—)6.5-12 X (1.5-)3-7.5 cm,
oblanceolate to broadly obovate or broadly elliptic,
brous;
glabrous, coriaceous, glossy, apex rounded to
broadly cuspidate, base cuneate to attenuate, ‹
-
r
rounded: margins glabrous, entire; secondary veins
(6 or 7) to 10 pairs per side, eucamptodromous,
adaxially inconspicuous, abaxially conspicuous:
domatia absent; stipules 10—14 mm long, ovate to
obovate, abaxially carinate, glabrous, free at the
deciduous.
base, Inflorescences 4 to 8 per axil;
heads 1.5-1.8 cm wide; inflorescence axes 2.8—5
cm long, terete or slightly flattened; bracts calyptra-
like. persistent; peduncles 0.8—2 cm long, sparsely
pubescent. Flowers 4-merous; calyx tubes ca. 1 mm
long, completely fused, lobes ca. 1 mm long, inside
and outside densely pubescent; corolla tubes ca. 4
X | mm, red, glabrous, lobes ca. 1 mm long, broad-
—
ly oblong, yellow-tinged, glabrous; anthers ca.
mm long. Styles 7-8 X 1-1.5 mm; stigmas globose:
ovary 2-carpellate; carpel coherent; ovule ] per loc-
Inf,
Тах Сс С €
ule. pendulous; placentae small.
0.5-1.7 em diam., rugose, with calyx remnants
barely evident; individual fruits with endocarp
hard, glossy; disks accrescent, rounded, deeply di-
vided; seed 1 per locule, strongly flattened, red;
seed-coat lineate.
Habitat and distribution. | Evergreen rainforests,
occasionally in riverbeds; Districts of Andapa, Mor-
amanga, and Farafangana (Fig. 7).
Common names. Molompangady keliravina.
Marotsaka, Valompangady, Valotro, and Valotsy.
Phenology. Flowering May to August; fruiting
September to February.
Discussion. Breonia decaryana can easily be
distinguished from the rest of the Breonia species
by its complanate terminal vegetative buds and 4
to 8 inflorescences per axil. This species was re-
moved by Ridsdale from Вано simply because it
has flattened terminal vegetative buds and partly
fused corolla tubes. 1 here include it in Breonia
because this species shares one morphological syn-
apomorphy with Breonia sensu Ridsdale (large ac-
crescent disks); additionally, they all have multiple
fruits. | endorsed Ridsdale’s decision on sinking B.
keliravina in B. decaryana because the former has
all the diagnostic features of the latter. The speci-
men Louvel 216 (P!) was one of the two cited by
Homolle (1938) in the protologue of B. keliravina.
Additional specimens — MADAGASCAR.
Antsiranana: District Anda Réserve Spéciale
d ig eit -Sud, Г Жалы Том SG 257 (MO,
TA masina: District Mor ман Canton Anala-
ао. s.n. (Р), Louvel 216 (P); Ambodi-
voasary, 12-B-R-172 (TEF); he 26804 SF (TEF)
28414 SF (TEF), 28445 SF (TEF).
—
6. Breonia fragifera Capuron ex Razafim., sp.
nov. TYPE: Madagascar. Antsiranana prov-
nod District Antalaha, Canton Ampanavoana,
Antsiramoranga, 6 Dec. 1954 (fl), 6809 RN
(holotype, TEF). Figures 2C and 9.
Haec species a congeneris disco nectarifero accrescente
permagno atque fructu pustulato distinguitur.
Shrubs, spreading shrubs, or trees, 5-15 m tall.
Bark rugose. Leafy stems terete, glabrous. Terminal
vegetative buds conical, 3-3.5 X ca. З mm, gla-
brous. Leaves persistent; petioles 25-70 X ca. 1.5
mm, terete, glabrous; blades 8-13 х 2.2-A(-6.5)
20 Annals of the
Missouri Botanical Garden
NYA
Pbi! O ДУ
Bi
ri
Figure 9. Breonia fragifera.—A. Fertile branch with mature infructescence. —B. Fertile branch with inflorescence.
—C. Seed, showing flattened lateral profile (left); dorsal view (right). —D. Seed-coat texture. —E. Mature infructesc-
ence. Е Median dissection through flower, showing velutinous calyx tube and unicarpellate ovary with single pen-
Volume 89, Number 1
Razafimandimbison 21
Revision of Breonia
cm, oblanceolate to ovate, glabrous, coriaceous or
membranaceous, glossy, apex cuspidate to acute,
yase cuneate; margin glabrous, entire; secondary
veins 7 to 9 pairs per side, eucamptodromous: dom-
atia absent; stipules ca. 3 X 2 mm, cymbiform, not
carinate, glabrous, free at the base, deciduous. In-
florescences solitary; heads 1.5-2 cm wide: inflo-
rescence axes 1.2-3.5 cm long, terete; bracts ca-
lyptra-like, deciduous; peduncles 2-3 mm long.
Flowers mostly 4-merous; calyx tubes ca. 0.5 mm
long, velutinous, lobes lanate; corolla tubes ca. 4
X l mm, white-tinged, glabrous; lobes ca. 1 mm
long, oblong, ciliate or glabrous; anthers ca. 0.5 mm
long: filaments ca. 0.1 mm long, glabrous. terete:
styles ca. 7 X 0.2 mm, glabrous; stigmas capitate:
ovary 2-carpellate; carpels coherent; ovules 1 per
locule, pendulous. Infructescences 1.6-2 cm diam..
pusticulate, with calyx remnant barely evident; in-
dividual fruits with endocarp soft, fibrous; disks ac-
crescent, massive, obtriangular; seed 1 per locule,
strongly flattened, ellipsoid, red; seed-coat lineate.
Habitat and distribution. | Low-altitude forests:
Districts of Ambanja, Antalaha, Fort-Carnot, Toa-
masina IL, Fort-Dauphin (Fig. 7).
Common names. Valotra, Valotralahy (male
Valotra).
Phenology. Flowering November—March; fruit-
ing June to March.
Discussion. This species is diagnosed by its
massive, accrescent disks and pusticulate fruits. It
was lumped by Ridsdale (1975) in Breonia sphaer-
antha. The species name was taken from the labels
of the herbarium specimens of Breonia fragifera re-
ceived from TEF. The epithet “fragifera” indicates
strawberries-bearing, referring to its inflorescences.
Paratypes. MADAGASCAR. Antsiranana: District
ABhanj n 251 (MO); Beampangibe, 2961
F (TEF): ara, е Gautier et al, 3288
a Mos а nton — i 7 SF (TEF);
Presqu'le d’Ambato, foret classée, — etal.
324 (MO); District Antalaha, Canton Ampahana, An-
dranomadio, 2/570 SF (ТЕЕ). Fianarantsoa: District
Fort- — Canton lfanirea, 19791 | SF (TEF), —
SF (TE — Ambodiriana 9045 RN (ТЕЕ);
virons E E Baie d'Antongil, Massif d'Antsirosiro, 87 34
z
3
=a
E
T
SF (TEF); Toamasina Il, Ampasimbe, Andranotsara, SFF
Mihatsara. — i. SF (T EF). Comtet 33536 SF
(TEF). Com aput pd Comtet 34413 SF (ТЕР),
Noyes et al. p . MO, P, TAN); Réserve Naturelle In-
tégrale de ия — Sv е el
al. 259 (TAN). t Fort-Dauphin, 5
(P. TEF). Урда кыд p (MO); Canton rate
sa, Antsako, 10803 SF (ТЕЕ); Isaka-Ivondro au bords de
la rivière Kovazaza, 11503 SF (TEF).
7. Breonia havilandiana Homolle, Bull. Soc.
Bot. Fr. 84: 464. 1937 [publ. 1938]. TYPE:
Madagascar. [Fianarantsoa province]. “bords
de l'Anosivola (Mangoro)." 700 m, Sep. 1911
(fl), Perrier de la Báthie 3904 (holotype, P!).
Trees, 10-20 m tall. Bark fissured longitudinally.
Leafy stems glabrous, lenticellate. Terminal vegeta-
15-19 3.5—4 mm,
Leaves deciduous; petioles 15-25 X ca. 2 mm, te-
rete, glabrous; blades 8-19(-21) х 5-9(-12) ст, ob-
ovate, always rippled or wavy when dry, glabrous,
tive buds conical, glabrous.
coriaceous, apex obtuse to rounded, base cuneate;
margins glabrous, entire; secondary veins 8 to 10
pairs per side, eucamptodromous, abaxially promi-
nent; ciliate-type domatia; stipules ca. 20 X 3 mm.
cymbiform, abaxially carinate, glabrous, free at the
base. deciduous. Inflorescence solitary, rarely 2 per
axil, heads ca. 2.3-2.9 em wide; bracts calyptra-like:
inflorescence axes 3—4 cm long, glabrous: peduncles
ca. 34 mm long, glabrous. Flowers 5-merous; calyx
tubes ca. 3.5 mm long, inside at the base velutinous,
toward the lobes pubescent, outside glabrous. a few
straight long hairs surrounding the base, lobes ob-
long. tomentose; corolla tubes ca. 5 X 0.6—0.7 mm,
glabrous, lobes 15-2.5 mm long, oblong: anthers
1.1-1.2 mm long; filaments ca. 0.1 mm long. gla-
styles 8.5-9
brous; stigmas capitate, shallowly bifid; ovary 2-car-
brous, flattened: ca. 0.2 mm, gla-
pellate; carpels syncarpous; ovules 2 to 4 per locule,
flattened, pendulous, imbricate; placentae flattened,
elongated. Infructescence 1.7-2.5 cm diam., with
well-developed calyx remnants; individual fruits with
endocarp soft, not glossy; disks accrescent, obconi-
cal; seeds 2 to 4 per locule, strongly flattened, ellip-
sold, red; seed-coat reticulate.
Habitat and distribution. Low- and mid-alti-
tude evergreen rainforests; Districts of Ifanadiana,
Befandriana-Nord, Moramanga, Fort-Carnot, and
Anosibe an'Ala (Fig. 5).
Common names. Molompangady, Molotr'angady
(lips of spade), Mamalifangady.
Phenology. Flowering October to November:
fruiting September to January.
Discussion. This species is distinctive in having
leaf blades that are undulate when dry.
Additional examined. MADAGASCAR
specimens
—
dulous ovule each (left); entire flower (right). —
edian dissection through infructescence revealing — disk
and uniovulate carpels (left); entire fruit (right). ^. E F from 6809 RN (TEF) and C-E, G from 79775 SF (TE
Annals of the
Missouri Botanical Garden
= louvelii
+ richardsonii
14° ж sphaerantha
A stipulata
Ж taolagnaroensis
© tayloriana
50°
ү Е
14°
$
18°
22°
200 Kilometers
46°
Figure 10.
B. tayloriana.
‘ianarantsoa: District Fort-Carnot, Canton Tolongoina,
Ankadilalana, 19295 SF (TEF); Forêt d’ Ambutohareaha.
7136 SF (TEF); District Hanadiana, lo — al
Park, Randrianasolo Rasabotsy 32 (MO), 9974 SF
(TEF), Бакайата Нов SG 389 (MO, TAN). оь
ga: District Befane dri iana- Nord, eb Matsondakana, Be-
Héserve Spéciale d' Anjanaharibe- Sud, Ravelona-
rivo € Rabesaonina 570 ). Toamasina: District
Anosibe an'Ala, Tsaratampona, Mangorobilika, 25526 SF
(TEF); District Moramanga, Canton Périnet, Sahamaloto,
6105 SF (TEF
&
lalona,
—
8. Breonia louvelii Homolle, Bull. Soc. Bot.
France 84: 461. 1937 [publ. 1938]. TYPE:
Madagascar. [Toamasina province], Analama-
zaotra, Sep. 1925 (fl), Louvel 125 (holotype, P!).
Medium-sized trees, 7-10 m tall. Bark with lon-
gitudinal fissures. Leafy stems rounded, glabrous,
lenticellate. Terminal vegetative buds conical, 5—
1-1.5 mm long, glabrous. Leaves persistent; pet-
mm long, adaxially canaliculate, gla-
ioles ca. 7—17
^
brous; blades ca. 4-9 X 1.5-3.5 em, oblanceolate
to elliptic, glabrous, coriaceous, glossy, apex acu-
minate to obtuse, base attenuate; margins glabrous,
entire, slightly involute when dry; secondary veins
10 to 11 pairs per side, eucamptodromous, adaxi-
50°
Distribution of Breonia louvelii, B. richardsonii, B. sphaerantha, B. stipulata, B. taolagnaroensis, and
ally inconspicuous, abaxially conspicuous; domatia
absent; stipules 5-9 X 1.5-1.8 mm, cymbiform, not
carinate, glabrous, free at the base, deciduous. In-
florescence solitary, heads 1.8-2 cm wide; inflores-
cence axes 2—4 cm long, flattened, slender, gla-
brous; bracts calyptra-like, deciduous; peduncles
ca. 2 mm long, glabrous. Flowers 5-merous or rare-
ly 6-merous; calyx tubes ca. 1 mm long, free, lobes
ca. 1 mm long. inside lanate, outside glabrous,
lobes densely pubescent, truncate to oblong: corolla
tubes 56 X
long. broadly oblong; anthers 1—1.5 mm long; fila-
ments 0.2-0.3 mm long; styles 7-8 X 1-1.5 mm;
stigmas capitate (shallowly bifid); ovary 2-carpel-
1-1.2 mm, glabrous, lobes 2-3 mm
ate; carpels syncarpous; ovule 1 per locule, pen-
dulous; placentae small. Infructescences not seen.
Habitat and distribution. | Evergreen rainforests;
District of Moramanga (Fig. 10).
Common names. | Hazomarotsaka (slender but
tall trees), Molompangady.
Phenology. Flowering September to October:
fruiting December to January.
Discussion. This species has a very restricted
geographical distribution and appears to be rare.
Volume 89, Number 1
Razafimandimbison
Revision of Breonia
Additional e MADAGASCAR.
Toamasina: District Moramanga, Canton Périnet, Antan-
on dm 7921 SF (TEF); "ire УЛ 7562 SF (TEF); Ré-
rve —— d'Analamazaotra, Razafimandimbison SG
358 (MO, TAN).
spec imens
9. Breonia lowryi Razafim., sp. nov. TYPE: Mad-
agascar. Fianarantsoa province: Ambalavao,
Andringitra Reserve, Lewis et al. 755 (holotype,
MO 05066234; isotypes, К, TAN). Figure 11.
iec species a congeneris loculis corollinis glabris,
ubo calycino intus glabro atque loculis 7- ad 9-ovulatis
distinguitur
Trees, 10—30 m tall. Bark rugose. Leafy stems
quadrangular, glabrous. Terminal vegetative buds
ca. 3 X 2 mm, glabrous. Leaves persistent, petioles
10-14 mm long, adaxially canaliculate, glabrous:
blades (5.5—)7.8-12 2.4—)3.4—6.4 cm,
glabrous, coriaceous, not glossy, apex broadly cus-
pidate, base attenuate; margins glabrous, entire:
secondary veins 6 to 9 pairs per side, eucampto-
dromous, abaxially conspicuous; pit-type domatia
in the axils of the secondary veins, pubescent; stip-
ules ca. 5 X 1-1.2 mm, cymbiform, not carinate,
glabrous, free at the base, deciduous. Inflorescenc-
es solitary, heads 2-2.3 cm wide; inflorescence axes
ca. 1.8-3 cm long, flattened, glabrous; bracts ca-
lyptra-like, deciduous; peduncles ca. 2 mm long,
glabrous. Flowers 5-merous; calyx tubes ca. 2 mm
lobes
obovate,
long. inside glabrous, outside puberulent,
0.5-0.8 mm long, oblong, inside glabrous, outside
velutinous; corolla tubes ca. 6 X 0.4 mm, reddish,
inside puberulous, outside glabrous, lobes 2-2.2
mm long, oblong, puberulous; margins glabrous;
anthers 0.5-0.8 mm long; filaments ca. 0.2 mm
long, flattened; styles « 0.6 mm; stigmas
elongate to capitate; ovary 2-carpellate;
syncarpous; ovules 7 to 9 per locule, pendulous;
placentae elongated. 1-1.5 ст
diam., with well-developed calyx remnants; indi-
vidual fruits berry-like; t, fibrous:
disks accrescent, rounded; seeds
with rudimentary
carpels
Infructescence
endocarp soft,
7 to 9 per locule,
strongly flattened, wings, red:
seed-coat reticulate.
Evergreen forests be-
Districts of Fort-
Habitat and distribution.
tween 900 and 1500 m altitude:
Carnot, Moramanga, Ambalavao, and Fianarantsoa
(Fig. 12).
Common names. Molompangady, Valompangady.
Phenology. Flowering time unknown: fruiting
May and October.
Discussion. This species differs from the other
Breonia species that also have 7 to 9 ovules per
locule and glabrous corolla lobes because the in-
side of the calyx tubes is glabrous. The epithet of
this new name honors Porter P. Lowry Il, head of
the Africa and Madagascar Department of the Mis-
souri Botanical Garden, who has done much to fur-
ther studies of the Araliaceae as well as other plant
families in Madagascar.
Paratypes. MADAGASCAR. енли District
Fianarantsoa Il, Canton Ampamaherana, 2078 SF (TEF):
District Fort-Carnot, To E aes ee
аз SF (TEF), 11594 SF (TEF), 14-B-R-230 (TEF
asina: District = йы Forêt d’ Fede
Périnet), 24150 SF (TEF).
10. Breonia macrocarpa Homolle, Bull. Soc.
Bot. Fr. 84: 461. 1937 [publ. 1938]. TYPE:
“Madagascar,” Pérrier de la Bathie 3933 (lec-
totype, designated by Ridsdale (1975: 545).
P!). Figures 2D and 3
Trees, 10—20 m tall. Bark longitudinally fissured.
eafy stems quadrangular, glabrous. Terminal veg-
etative buds conical, 2740 X 4-6 mm, puberu-
lent. Leaves persistent; petioles 35-60 X ca. 4 mm,
terete, lenticellate, puberulous; blades 19.5-38 x
15-25 em, broadly ovate to broadly oblanceolate.
brown-tinged when dry, adaxially glabrous, abaxi-
ally sericeous, coriaceous, glossy, apex cuspidate to
rounded, base rounded; margins glabrous, entire:
secondary veins 12 or 13 pairs per side, eucamp-
todromous, abaxially densely pubescent, adaxially
glabrous: tuft-domatia in the axils of secondary and
tertiary veins; stipules 27-34 X 6-9.5 mm, cym-
biform, abaxially carinate, puberulous, free at the
Inflorescences solitary, heads ca.
base, deciduous.
4.7 cm wide; inflorescence axes 8.5-12.5 cm, flat-
tened, glabrous to puberulous; bracts calyptra-like.
deciduous; peduncles not elongated. Flowers 5-
merous or rarely 4-merous; calyx tubes 2.8—3 mm
long, green, below the middle velutinous, around
the mid-part glabrous, toward the lobes puberulous:
lobes ca. 2 mm long, truncate to oblong, velutinous:
corolla tubes ca. 11 mm long, glabrous, lobes ca.
3.5 X 0.25 mm long, oblong, glabrous. alado
toward the apices; margins ciliate; anthers ca.
mm long; filaments 0,5—1 mm long, flattened. Styles
21 X 0.5 mm, glabrous; stigmas clavate; ovary
2-carpellate; carpels syncarpous: ovules 4 to 6 per
locule, flattened, pendulous, imbricate; placentae
strongly flattened, elongated, adnate to the septum.
Infructescences 3—4 cm diam., woody when dry,
with well-developed calyx remnants;
fruits with endocarp hard, glossy; disks accrescent,
pentagonal; seeds 4 to 6 per locule, strongly flat-
tened, ellipsoid, red, reticulate.
individual
Phenology. Flowering March to May, August to
24
Annals of the
Missouri Botanical Garden
September; fruiting May to July, December to Jan-
uary.
Habitat and distribution.
tude of eastern evergreen rainforests; Districts of
Brickaville, Moramanga, and Toamasina П (Fig. 5).
Common names. рар заду апа Valotra.
Discussion. Ridsdale (1975: 545) included
Breonia macrocarpa as a synonym of В. madagas-
Low- and mid-alti-
cariensis, dismissing the former as "a hairy-leaved
form" of the latter. However, these two are quite
different morphologically, as is clear from Table 3.
Despite the striking morphological differences be-
tween the two species, B. macrocarpa has not been
recognized since its original description, partly be-
cause Homolle (1938) did not provide kevs for the
species of Breonia she recognized and partly be-
cause no collections of B. madagascariensis have
been made for almost 200 vears.
While describing this species, Homolle cited
three specimens, Thouvenot 117 (P!), Perrier de
Báthie 3933 (P!), and Perrier de la Báthie 5270
(P!), without selecting the type species.
Additional examined. MADAGASCAR.
Toamasina: District Brickaville, Fetromby, Ambalakon-
3265 SF (TEF); District Moramanga, Perinét, 26906
F (TAN), Noe el i Mg K, MO, P, TAN), Razafiman-
pd төз SG 352 Р, TAN): District Tenia П.
Canton Ambodiriana, аад, 9052 RN (Р), 91:
RN (Р); Canton Ambodiriana, SF (P, TEF), 21160
SF (TEF), 28136 SF (TEF), 32344 SF (TEF).
specimens
m
N
©
11. Breonia madagascariensis A. Rich., in
DC., Prodr. 4: 620. Sep. 1830. Sarcocephalus
madagascariensis (A. Rich.) Baill., Adansonia
12: 311. 1879. TYPE: “Madagascar.” [Without
exact locality], Chapelier s.n. (lectotype, des-
ignated by Ridsdale (1975: 545), K!;
totype, Р!). Figure ТЕ.
isolec-
Trees, height unknown. Bark rugose. Leafy stems
quadrangular, glabrous. Terminal vegetative buds
conical, ca. 35 X 6 mm, glabrous. Leaves persistent,
x 35
glabrous,
subsessile; blades ca. 45 cm, broadly obovate
to broadly spathulate, coriaceous, not
glossy, apex caudate, base attenuate; midrib angular,
lenticellate, prominent; margins glabrous, entire;
secondary veins 18 or 19 pairs per side, eucamp-
todromous, abaxially prominent; domatia absent;
stipules ca. 35 mm long, cymbiform, abaxially car-
inate, glabrous, free at the base, deciduous. Inflo-
rescences solitary, heads ca. 6 ст wide; inflores-
cence axes 16-21 cm long, flattened, puberulous;
bracts calyptra-like, deciduous; peduncles not elon-
gated. Flowers 5-merous; calyx tubes ca. З mm long,
with prominulous ribs, inside lanate, outside gla-
brous except on upper parts of ribs, lobes ca. 2 mm
long. oblong to truncate, pubescent; corolla tubes c:
З х 1
mm, oblong, inside glabrous, outside puberulous in
1.5 mm, ribbed, glabrous, lobes са.
upper third, margins ciliate; anthers ca. 2 mm long;
filaments 0.25—0.5 mm long, flattened; styles 17-18
X са. 0.25 mm, glabrous; stigmas clavate; ovary 2-
carpellate; carpels syncarpous; ovules 8 per locule,
pendulous, imbricate; placentae strongly flattened,
elongated. Infructescences not seen.
Habitat and distribution, Habitat unknown:
Madagascar.
Phenology. Unknown.
Discussion. Breonia madagascariensis is known
only from the six authentic herbarium specimens
collected by Commerson, Chapelier, and Bréon
during the early botanical explorations they under-
took separately in 1770-1771, 1794-1808, and 5
September-5 December 1818, respectively (Dorr,
1997). No collecting localities were mentioned on
their labels. This species has not been collected
since 18
Chapelier is known to have resided on the East
Coast at Foulpointe for a dozen years, and he es-
tablished an experimental garden in a place known
locally as Isatrano (Tamatave or Toamasina). Also,
he made botanical collections around Ivoloina,
Foulpointe, and on the island of St. Marie (all with-
in Toamasina province). Hence, it is possible that
Breonia madagascariensis was collected from the
above-mentioned areas of the east coast of Mada-
gascar. Most of the eastern coast of the country has
been deforested for agricultural purposes and re-
placed by secondary vegetation. characterized. by
the dominance of Ravenala madagascariensis Sonn.
(Strelitziaceae). However, B. madagascariensis may
still be extant inside protected forests (Betampona
Reserve, Zahamena Reserve, Analamazaotra Re-
serve) in this area, so efforts to rediscover В. mad-
agascariensis should focus on these three protected
areas and their surroundings.
=>
Figure 11. Breonia lowryi.—A. Fertile brane i with TUE infructescence. —B. Mature infructescence. С. Seed,
showing tate ne E — ral profils (left); dorsal view (right). —D. Median dissection through fruit, showing calyx tube and
bicarpellat ith 7 to 9 pendulous ovules th r locule ic lo entire fruit (right). —E. Two adjacent flowers, showing
two dH. ovaries separated: median dissection through flower, showing glabrous corolla amd pubescent calyx tubes
and bicarpellate ovary with ovules (left); entire flower, showing only part of corolla tube (right).
). Lewis et al. 755 (MO)
Volume 89, Number 1
2002
Razafimandimbison
Revision of Breonia
Annals of the
Missouri Botanical Garden
Table 3.
separating Breonia macrocarpa and В. madagascariensis.
Distinctive vegetative morphological features
Breonia В. madaga 5-
Characters macrocarpa cariensis
Petioles well developed subsessile
Leaf blade size 19-38 X 15-25 cm 45 X 35 cm
Abaxial indumentum densely pubescent glabrous
on leaf blade
Midrib shape rounded angular
Domatia present absent
Peduncle length 8.5-12.5 cm 16-21 cm
The lectotype Chapelier s.n. selected by Ridsdale
(1975) was from the six known collections.
Additional sj ns examined. “Madagascar, Herbier
Drake, P- 001. 3910 (Py: "Madagascar, herbarium Ri-
d P-001 а (P). Bréon s.n. (P). Commerson s.n.
(P not seen).
J. Linn. Soc.
"Madagascar.
12. Breonia membranacea Havil..
Bot. 33: 36. 1897. TYPE:
[Without exact locality], Perottet s.n. (lecto-
type, designated by Ridsdale (1975: 545), Р!).
Trees, 12-18 m tall. Bark rugose. Leafy stems te-
rete, glabrous. Terminal vegetative buds conical, 10—
20 X ca. 1.5 mm, puberulent at the base. Leaves
persistent; petioles 3-6 mm long, adaxially canalicu-
16, puberulous when young, glabrescent; blades 8.5—
{ 4.5-5.4 ст. oblong, glabrous, membrana-
ceous, adaxially glossy, apex broadly cuspidate to
acute, base rounded; margins glabrous, entire; sec-
ondary veins ca. 8 pairs per side, eucamptodromous,
adaxially prominulous; domatia hairy in the axils of
the secondary veins; stipules 12-21 mm long. cym-
biform,
uous. Inflorescences solitary, heads ca. 2.2 cm wide:
not carinate, glabrous, free at the base, decid-
inflorescence axes 4.5—5.5 em long, strongly flattened.
glabrous; bracts calyptra-like, deciduous; peduncles
3 mm long. glabrous. Flowers 5-merous: calyx
tubes 1—1.2 mm long, inside lanate, outside glabrous.
lobes 0.5-0.7 mm long. oblong, velutinous: corolla
tubes ca. 5 X 0.50.7 mm, inside puberulous, outside
glabrous, lobes ca. 2 mm long, obtuse to truncate,
puberulous, margins ciliate: anthers ca. 1 mm long:
filaments ca. 2 mm long, flattened, glabrous; styles 7—
8 X ca. 0.5 mm, glabrous; stigmas capitate; ovary 2-
carpellate; carpels syncarpous; ovules З per locule.
flattened, pendulous, imbricate; placentae small, elon-
gated. Infructescences 1.5-1.8 cm diam., with well-
developed calyx remnants; individual fruits with en-
docarp soft, fibrous; disks accrescent, heart-shaped:
seeds 1 to 3 per locule, flattened, concavo-convex,
ellipsoid, red; seed-coat reticulate.
Habitat and distribution. Evergreen lowland
rainforests, littoral forests; Districts of Moramanga
and Fénérive Est (Fig. 12).
Common names. Molompangady, Valotra.
Phenology. Flowering January February;
fruiting March to April.
Discussion. This species is distinct in its leaves
with relatively short petioles. The lectotypification
made by Ridsdale is questionable because the lec-
totype, Perottet s.n., does not match with Haviland's
(1897) remarks concerning the terminal vegetative
buds of B. membranacea: “| have seen none [spec-
imens] with the apex uninjured." Perrottet s.n. has
an intact and uninjured apex and was annotated by
Haviland as Breonia membranacea. Haviland clear-
ly stated that the type specimen of B. membranacea
was in the Paris herbarium. Haviland mentioned
that the description of Breonia mauritiana was
based on a single specimen: he did not mention
that this was also the case for B. membranacea.
Hence, he could possibly have seen more than one
specimen of B. membranacea, although he did not
cite any other specimens in the original description.
Haviland is known to have visited the herbaria
at the British Museum (BM), Leiden (L), and Paris
(P) (Haviland, 1897). However, from Kew (K). L.
and P, I received only the specimen of Perrottet s.n.
from P. It is possible that the "injured" specimens
seen by Haviland are at BM. Ridsdale visited all
three herbaria while preparing his revision of Af-
rican and Madagascar Naucleeae (Ridsdale, 1975);
he did not cite any syntype for B. membranacea.
This implies that he saw only the specimen of Per-
rottet s.n. at Paris. Before one can attempt to resolve
the nomenclatural problem of B. membranacea, one
must visit BM, L, and P and look for possible spec-
imens of B. membranacea annotated by Haviland.
Additional specimens о ed. MADAGASCAR
Toamasina: District. Fénérive-Est — Tampolo,
13094 SF (TEF), 16496 SF (TEF). 16621 SF (P); District
Moramanga, Forêt d'Analamazaotra, Louvel 186 (P).
13. Breonia perrieri Homolle, Bull. Soc. Bot.
Fr. 84: 461. 1937 [publ. 1938]. TYPE: Mad-
agascar. [Toliara province]: Morondava, Forét
3513
designated by Ridsdale (1975:
de Marofandilia, Perrier de la Báthie
(lectotype,
)
546),
Medium-sized trees, ca. 8 to 15 m tall. Bark with
longitudinal fissures. Leafy stems flattened, lenticel-
ate, glabrous. Terminal vegetative buds conical, 9—
11 X 34 mm, glabrous. Leaves deciduous: petioles
33-60 X 1.5-1.8 mm, terete, glabrous or pubescent
toward the apex, lenticellate; blades 9-15 X 6-1:
Volume 89, Number 1
2002
Razafimandimbison 27
Revision of Breonia
50°
N
[=] lowryi He:
14° Ш membranacea d 14?
@ sambiranensis
* tsaratananensis
18° 18°
22? 229
200 Kilometers
46^
Figure 12.
cm, broadly obovate to orbicular, pubescent or gla-
brous, coriaceous or rarely membranaceous, apex
cuspidate to rounded, base cordate; margins gla-
brous, entire; secondary veins ca. 9 pairs per side,
the first 3. diverging from the base of the midrib,
prominulous; domatia absent; stipules 8-14. X 34.2
mm, cymbiform, abaxially carinate, glabrous, free at
the base, deciduous. Inflorescences solitary, heads
2.5-2.0 cm wide; 2.4-3.5 cm
long, terete, pubescent or glabrous; bracts calyptra-
inflorescence axes
ike, deciduous; peduncles 2—3 mm long. Flowers 5-
merous; calyx tubes ca. 2.5 mm long, green, inside
below the middle glabrous, velutinous from the mid-
dle toward the base of lobes, outside glabrous, lobes
1—1.2 mm long, oblong, puberulous; corolla tubes 4—
5 X ca. 1 mm, yellow-tinged, glabrous, lobes ca. 1.5
mm long, oblong, glabrous; anthers 1-1.1 mm long;
filaments ca. 1 mm long, flattened, glabrous; styles
х 0.2
vate; ovary 2-carpellate; carpels syncarpous; ovules
mm, glabrous; stigmas globose to cla-
2 to 3 per locule, strongly flattened, pendulous, im-
bricate; placentae small, elongated. Infructescences
8-2.2 cm diam., with well-developed calyx rem-
nants; individual fruits with endocarp hard, glossy;
disks accrescent, heart-shaped; seeds 1 to 2 per loc-
50°
Distribution of Breonia lowryi, B. membranacea, B. sambiranensis, and B. tsaratananensis.
ule, with rudimentary wings at the base, strongly flat-
tened, ellipsoid, white-tinged; seed-coat reticulate.
Phenology. Flowering November January;
fruiting January to March.
Habitat and distribution. Western and. north-
western deciduous dry forests; from Districts of
Antsiranana П and Ambato-Boeni, Befandriana-
Nord,
tirano, Marovoay,
Antsalova, Belo Tsiribihina, Main-
Ambilobe, and Vohémar (Fig. 5).
otra.
Soalala,
Common name. Val
Phenology.
fruiting January—February.
Flowering November—December;
Discussion. The following were among the four
specimens cited by Homolle in the protologue of
Breonia perrieri: Perrier de la Báthie 825 (Р not
Perrier de la Báthie 17852 (P not seen
Greve s.n. (P not seen).
. and
м
seen),
Additional specimens examined. MADAGASCAR, An-
tsiranana: District Antsiranana 5 Ankarana, Ambiloman-
godro, Forêt de — 29653 SF (TEF); eta Ré-
serve Spéciale d'Ankarana, Malcomber et al. 1833 (K, MO,
P, TAN), Rasafimandirdicon SG 273 QB 2s Forét
a Ambis, 6181 SF (TEF), 6616 SF (TE Е); Plateau
d'Ankarana à l'ouest d'Ambondromifehy, 3026 SF (K, TEF);
District Vohémar, Forét d’Analafiana au nord de la base
Annals of the
Missouri Botanical Garden
Manambery, 27503 SF (TEF); 14 km E of Vohémar, near
Analafiana, 15673 SF (TEF); Ankara, 6696 SF (TEF). Ma-
hajanga: District Ambato-Boeni, Ankirihitra, Forêt
d'Anatialabe, 19379 SF (ТЕЕ); Di: MN Befandriana-Nord,
Ambahivahy, Forêt d'Andembikely, 79061 SF (TEF); кон êt
de Marohogo à l'ouest de — 18450 SF (TEF): Di
trict — a 3 km à FF Bevitika, Labat & Conté
2674 (K, › TAN), — et * 2261 (К. MO, P. TAN):
Plateau de zh ко aux environs de — 6754 5
(TEF), Jongkind et a 3412 (K, MO. P, ). Villiers et al.
4846 (MO. P, TAN); District Soalala, че, Andrano-
2 SF PUN Andranomavo, 17565 SF (TEF). To-
B — Belo d Tsimafana, Ankirijifotsy,
7987 SF (TEF), Noyes et al. 1028 (K, MO, P, TAN); District
те lava, Tsimembo-limite concession de M. Barthe,
8245 SF (TEF); Ankilatsy, 4677 SF (TEF)
14. Breonia richardsonii Razafim., sp. nov.
TYPE: Madagascar. [Toamasina province]: Ma-
roantsetra, Jardin botanique de Farankaraina,
16 June 1965 (fl), 14.359 SF (holotype, TEF).
Figure 13.
—
me
aec species a congeneris inflorescentiae bracteis tub-
ulosis adpressis in | triangulares tres quatuorve desinen-
tibus atque ovariis contiguis post anthesin distinguitur.
Trees, 15-30 m tall. Bark fissured longitudinally.
Leafy stems terete, glabrous. Terminal vegetative
buds conical, 34 X 1-1.5 mm, glabrous. Leaves
persistent; petioles 7-10 X ca. | mm, terete, gla-
brous: blades 5-7.7 X 2.5-3.7 em, oblanceolate to
ovate, glabrous, membranaceous, apex cuspidate to
acute, base acute; margins glabrous, entire; second-
ary veins 5 to 7 pairs per side, eucamptodromous,
inconspicuous; domatia cryptic-type in the second-
ary veins, glabrous, adaxially prominent; stipules ca.
5 X 1-1.5 mm, eymbiform, not carinate, free at the
base, deciduous. Inflorescence solitary, heads ca. 2
cm wide; inflorescence axes 2.5—3 cm long, terete,
glabrous: peduncles ca. 5 mm long; bracts tubular
with 3—4 broadly triangular lobes, not enclosing the
young inflorescence, persistent. Flowers 5-merous or
rarely 4-merous; calyx tubes ca. 1 mm long, inside
glabrous, except at the base velutinous, outside at
the base velutinous, in the mid-parts glabrous, to-
ward the lobes puberulous, lobes broadly triangular,
0.3-0.5 mm, gla-
1.5 mm long, oblong, glabrous: an-
puberulous; corolla tubes са. 4 X
brous, lobes Ca.
thers 2-2.5 mm long: filaments ca. 0.1 mm long.
flattened: styles ca. 7 X 0.5 mm, glabrous; stigmas
globose to clavate: ovary 2-carpellate, adjacent ova-
ries fused only at the base; carpels syncarpous;
ovules 5 per locule, pendulous, imbricate: placentae
large, elongated, occupying the upper third of the
locule. Infructescences 1—1.5 cm diam., with well-
developed calyx remnants, adjacent ovaries fused up
to the mid-point; individual fruits with endocarp soft,
fibrous; disks accrescent, obconical; seeds 5 per loc-
ule, strongly flattened, ellipsoid, angular, red; seed-
coat reticulate.
Habitat and distribution. Lowland rainforests;
District of Maroantsetra (Fig. 10
Common names. — Valotrafotsy.
Phenology. Flowering April; fruiting November.
Discussion. This species is very different from
other species of Breonia because of its tubular, ap-
pressed inflorescence bracts with 3 to 4 broadly
triangular lobes and post-anthesis fusion of adja-
The specific. epithet honors. Mick
cent ovaries.
Richardson, who was my Ph.D. advisor as well as
the Manager of the Graduate Program at the Mis-
souri Botanical Garden. This species is extremely
rare and has not been collected since 1965.
Paratype. MADAGASCAR. Toamasina: District Ma-
roantsetra, Jardin botanique de Farankaraina, 18311 SF
(TEF)
15. Breonia sambiranensis Razafim., sp. nov.
TYPE: Madagascar. [Antsiranana province]:
Nosy be, Réserve Naturelle Integrale de Lo-
kobe, 13°24'S, 48?20'E, 350 m. 7 July 1995
(fl), Antilahimena 237 (holotype, TAN; isotype,
MO). Figures 3C, 4D, and 14.
species a congeneris calycis tubo brevi аши
costato et lobulis brevibus in centro non profunde foveo
latis,
directas gerentibus distinguitur
ad margines protuberationes breves stylum versus
(6-10-20 m tall. Bark
smooth, glabrous. Leafy stems quadrangular, gla-
Trees, rarely shrubs,
brous, lenticellate. Terminal vegetative buds coni-
cal, ca. 5 X 2-3 mm, glabrous. Leaves persistent;
үк 15-30 mm long, adaxially canaliculate,
glabrous; blades (8.5-)]2-17.5 X 4.7-9.5 cm,
broadly ovate to broadly obovate, or broadly ovate
to narrowly ovate, glabrous, coriaceous, glossy, apex
mucronulate to obtuse,
mucronulate to acute or
-
base cuneate, lower surfaces vellow-red-tinged
when dry; margins entire, glabrous; secondary veins
5 to 9 per side, eucamptodromous, dark red when
dry, abaxially prominulous; domatia absent; stip-
ules 6-7 X 4-5 mm, cymbiform, not carinate, gla-
the base, deciduous. Inflorescence
brous, free at
solitary, heads 3-3.7 em wide: inflorescence axes
2.5-3 cm long, flattened, glabrous: bracts calyptra-
like, deciduous; peduncles 1.5-2.2 mm long, gla-
brous. Flowers 5-merous; calyx tubes 1-1.2 mm
long, infundibular, prominulously ribbed, glabrous;
lobes 0.2-0.3 mm long, truncate, bearing a shallow
depression in the center, a short protuberance on
the edge toward the style (Fig. 14D). tomentose;
corolla tubes 7-8 X 0.8-0.9 mm, inside puberu-
lous, outside glabrous, lobes ca. 3 mm long, oblong,
—
Volume 89, Number 1
2002
Razafimandimbison 29
Revision of Breonia
recurved, with protuberance abaxially (Fig. 14E),
glabrous, with a few scattered hairs at the apex,
ciliate; anthers ca. 1.2 mm long; filaments 0.1—0.2
mm long; styles ca. 13 X 0.1 mm; stigmas clavate:
ovary 2-carpellate; carpels syncarpous; ovules 4 to
5 per locule, flattened, imbricate: placentae elon-
gated, strongly flattened. Infructescences 1.2-1.6
cm diam., with accrescent calyx remnants; individ-
ual fruits with endocarps soft, fibrous; disks ac-
crescent, napiform; seeds 4 to 5 per locule, con-
cavo-convex, ellipsoid, red; seed-coat reticulate.
Habitat and distribution. This species occurs
in low- and mid-altitude evergreen rainforests; Dis-
tricts of Nosy Be, Ambanja, and Maroantsetra (Fig.
12)
Common name. Valotra
Phenology. Flowering March to July; fruiting
October to November.
This
lyx tubes aceite ribbed, short calyx lobes with
Discussion. species is distinct in its short ca-
shallow depressions in the center, and short protu-
berances on the edge toward the style (Fig. 14D). The
specific epithet refers to the Sambirano regions.
Paratypes. MADAGASCAR. Antananarivo: District
Ankazobe, Forét d'Ambohitantely, 84/3 SF (TEF). Antsi-
ranana: District Ambanja, Manongarivo Special Reserve,
* Bekolosy, 71430 ; District Nosy be,
SF (ТЕР), E 407 SF (TEF). 24753 SF
(TEF), 5524 RN (TEF), Birkinshaw 59 (K, MO); Bemarivo,
Androranga, ravin be Betsomanga, 841 SF (K, MO, ТЕР).
Toamasina: District Maroantsetra, Masoala Peninsula, S
of the village of pear pesmi Behasy & Vasey 203 (MO),
Rabe 115 (MO)
16. Breonia sphaerantha (Baill.) Homolle ex
Ridsdale, Blumea 22: 546. 1975. Franchetia
sphaerantha Baill., Bull. Soc. Paris 60:
477. 1885. TYPE: Madagascar. [Without exact
locality], Hildebrandt 3309 (holotype, P!; iso-
types, K!, 0).
Linn.
"р fongipetiovaiism Soler., Bull. Herb. Boissier
. 1893. TYPE: Madagascar. [Without exact lo-
ics Hildebrandt 3309 (holotype, L!; isotypes, K!
P!). Figure
Breonia — Вай, J. Linn. Soc. Bot. 33: 37. 1897.
TYPE: “Madagascar.” [Without exact loc v Hil-
debrandt pere (holotype, K*; isotypes, L!,
—
Shrubs to large trees, 7-18 m tall. Bark white-
tinged, rugose. Leafy stems terete, puberulous to
pubescent, lenticellate. Terminal vegetative buds,
2-3
persistent; petioles 15-20 X 0.9-1 mm, terete, pu-
26.7 X 2.5-3.9 cm, elliptic to
oblanceolate (oblong to oblanceolate), adaxially gla-
conical, 1.2-1.5 mm, puberulous. Leaves
berulous: blades 5
brous, abaxially puberulous, membranaceous,
glossy, apex acuminate, base rounded to truncate;
margins glabrous, entire; secondary veins ca. 7
pairs per side, eucamptodromous, barely evident on
upper surfaces, red-tinged when dry; domatia pock-
et-type in the axils of secondary veins, glabrous:
stipules ca. 5 X 2 mm, cymbiform, not carinate,
puberulous, free at the base, deciduous. Inflores-
cence solitary, heads 1—1.2 cm wide; inflorescence
axes 2.4-3.8 cm long, usually bent and twisted
when dry, terete, pubescent to velutinous; bracts
calyptra-like, deciduous; peduncles ca. 1 mm long.
Flowers mostly 4-merous or rarely 5-merous; calyx
tubes ca. 1 mm long, green-yellow-tinged, inside
velutinous, outside pubescent, lobes ca. 0.5 mm
long, triangular, pubescent; corolla tubes 4.2—4.3
ca. 0.2 mm, yellow-tinged, glabrous, lobes ca.
1.2 mm long, elliptic, lower parts glabrous, toward
the apex puberulous, ciliate; anthers ca. 1 mm long:
filaments ca. 0.2 mm long, terete, glabrous: styles
0-7 X ca. 0.8 mm, glabrous; stigmas globose to
clavate: ovary 2-carpellate; carpels coherent; ovule
| per locule, pendulous; placentae small, ovate. In-
fructescences 4-10 mm diam., densely pubescent,
with well-developed calyx remnants: individual
fruits with endocarp hard, glossy: disks accrescent,
obconical: seeds 1 per locule, with one side convex,
the other flattened, ellipsoid, red; seed-coat lineate.
Lowland evergreen
Nosy Be,
Ambanja, and Befandriana-Nord (Fig. 10).
Habitat and distribution.
rainforests; Districts of Antsiranana II,
Phenology. Flowering January—July; fruiting
March—December.
Baillon (1885), Solereder (1893),
and Haviland (1897) all used different sheets of the
same collection, Hildebrandt 3309, as the basis for
Franchetia sphaerantha,
Discussion.
Elattospermum longipe-
tiolatum, and Breonia parvifolia, respectively. Nei-
ther Solereder nor Haviland knew about Franchetia
sphaerantha, and Haviland did not cite Elattosper-
mum longipetiolatum either.
Idditional specimens examined. MADAGASCAR.
Antsiranana: Dist dg Antsiranana Il, Canton Anamakia,
Ankotikona, 15273 SF (TEF); bords de rivière Kongony,
Ambaliha. 2950 pes (T " TEF): District Nosy be, Lokobe
Reserve, Antilahimena 62 (K, MO); Lokobe, 7822 RN
(TAN). 0229 RN (TAN), 9453 RN (TEF), 11417 SF (TEF),
24764 SF (TEF); District Ambanja, 5б уйрет 255
(МО); — Marovato, Beangona, Rakoto 9 (TAN). Ma-
— District Befandriana-Nord, Commune Tsaraho-
папа, Forêt domaniale d'Antetezana, 15947 SF (TEF);
Se ee locality], 6257 RN (ТЕЕ).
=
17. Breonia stipulata Havil., J. Linn. Soc. Bot.
33: 35. 1897. TYPE: “Madagascar, Northwest”
[without exact locality], 1841 (fr), Pervillé s.n.
(holotype, P!).
30 Annals of the
Missouri Botanical Garden
F
uim m H
E
E i | E
"| ARE 2
1 (0 =
o f
Figure 13. Breonia richardsonii. —A. Fertile branch with mature infructescences. —B. Fertile branch with inflo-
ature infructescence, bearing a tubular bract. —D. Seed surface. —E. Seed, lateral view (left); dorsal
rescence. (С.
Volume 89, Number 1 Razafimandimbison 31
2002 Revision of Breonia
Trees or shrubs, 15-25 m tall. Bark rugose. Leafy 18. Breonia taolagnaroensis Razafim., sp. nov.
stems quadrangular, glabrous. Terminal vegetative
19-22 х 2-3 mn, glabrous. Leaves
deciduous; petioles 12-22 X 1.5-2 mm, adaxially
canaliculate, glabrous; blades 10—18.5 (224.5) X 4.5—
6(-8) cm, elliptic to lanceolate, glabrous, coriaceous,
buds conical,
not glossy, both apex and base acute; margins gla-
brous, entire; secondary veins (9—)11 or 12 pairs per
side, eucamptodromous, abaxially prominulous; dom-
atia absent (hairy or crypt-type domatia in the axils
of secondary veins); stipules 15-18 x 2.2-2.5(-3)
mm, cymbiform, narrowly carinate abaxially, glabrous,
ree at the base, deciduous. Inflorescences solitary,
heads 2.2-2.5 cm wide; inflorescence axes 4—6 cm
long, strongly flattened, glabrous: bracts calyptra-like.
deciduous; peduncles 1 mm long. Flowers 5-merous
or rarely 4-merous; calyx tubes ca. 2 mm long, inside
lanate, outside glabrous, toward the lobes puberulous,
lobes ca. 1 mm long, oblong, tomentose; corolla tubes
ca. 5 X 0.9-1 mm, glabrous, lobes 1.5-2 mm long,
oblong, glabrous; anthers ca. 1.5 mm long; filaments
0.2-0.3 mm long; styles ca. 10 X 1 mm,
stigmas clavate to slightly cylindrical: ovary 2-carpel-
glabrous;
late; carpels syncarpous; ovules 1 to 2 per locule,
pendulous, imbricate; placentae ovate, small. Infruc-
tescences 1.5-1.7 cm diam., with well-developed ca-
lyx remnants; individual fruits with endocarp hard,
glossy; disks accrescent, heart-shaped; seeds 1 per
locule, pendulous, concavo-convex, ellipsoid, both
ends acute, red; seed-coat reticulate.
Habitat and distribution.
Districts of Maintirano, Sambava,
Antsiranana II, and Antsalova (Fig. 1C
Deciduous dry for-
ests: Andapa,
Common names. — Valitsy and ИОНИ НОРРРРЕ И
Phenology. Flowering October to December:
fruiting December to February.
Additional specimens examined. MADAGASCAR. An-
tsiranana: District Ds Mandena, Forét de — jy
— ¢ паа 2 9 (TAN): lisière de forêt,
otozafy & Ral — 22 a т AN
Marojejy Est. f Mandena,
TAN); District Asana II,
tamisakana, 15057 SF
lova, | km à l'Est f
P); Botomena, 11089 RN (TEF); Tsiandro, Forêt d'Antsingy
aux environs de la clairière de piste d’Ambodiriana, 6791
); trail to the summit | of
Miller 3609 AS MO, P,
1 ‚ Ambarara-
TYPE: Madagascar. [Without exact locality].
Commerson s.n. (holotype, P!). Figure 2F.
Haec species a congeneris disco nectarifero pentagono
accrescente permagno distinguitur; ex sylvis littoralibus
prope Taolagnaro tantum cognita.
Shrubs, 4—6 m tall. Bark rugose with annular
fissures. Leafy stems quadrangular, glabrous. Ter-
) X
minal vegetative buds conical, 5 а. 3 mm,
^
glabrous. Leaves persistent; petioles 13-20 X ca.
2.5 mm, adaxially canaliculate, glabrous; blades 9—
11(-16) X 3.5-6.5 em, ovate to oblong, glabrous,
coriaceous, glossy, apex rounded to broadly cuspi-
date, base cuneate; margins glabrous, entire; sec-
ondary veins (7—)9 to 10 pairs per side. eucamp-
todromous; domatia absent; stipules 6-10 mm long.
cymbiform, not carinate, free at the base, glabrous,
deciduous. inflorescence
axes (4—)6—8.5 cm long, strongly flattened: bracts
not seen; peduncles absent. Flowers not seen. In-
Inflorescence nol seen;
fructescences 1.5-2 em diam., with well-developed
calyx remnants; individual fruits with endocarp
hard, glossy; disks accrescent, pentagonal; seed 1
per locule, concavo-convex, with rudimentary wings
at both ends, red; seed-coat reticulate.
and
evergreen rainforests; Districts of Fort-Dauphin and
Farafangana (Fig. 10).
Common names.
and Marotsaka.
Habitat and distribution. Littoral forests
Valotr'angady, Molompangady,
Phenology. Flowering unknown: fruiting Octo-
ber to December.
Bosser (1984) argued that the type
specimen of Cephalidium citrifolium (Poir.) A.
Rich. is the Commerson collection on which I base
Discussion.
Breonia taolagnaroensis. | disagree with Bosser be-
cause Commerson s.n. now in P has fruits only, but
Richard used a specimen with both flower and fruit
on which to base his Cephalidium. Commerson s.n.
has a flower only and matches Poiret’s original de-
scriptions of his Nauclea citrifolia.
Breonia taolagnaroensis is distinguished from the
other species of Breonia by its massive, accrescent
pentagonal disks, and is known only from the littoral
SF (TEF), 12023 SF (TEF), т? з G EF): Forêt de Va- forests of the Taolagnaro (Fort-Dauphin) regions,
loala, [collector unknown], 2200 (T: thus the specific epithet. The type specimen of this
<
view (right). —F. Median dissection — flower, showing bicarpellate ovary with numerous ovules (left): entire flower,
showing part of corolla tube
entire individual fruit (right).
3s d carpel, showing
. Dissection through a cay x remnant and a carp
el of a separate fruit (left):
la ent disk adnate to the septum and 5 ovules
БЕ).
=
=
attached to the placentae. A-F бой 14359 SF (TEF) and G from 18311 SF (TE
32 Annals of the
Missouri Botanical Garden
\
en KM MP
= 00). „
и IFA
\. Fertile branch with infructescences. —B. Fertile branch with inflorescences
—К. A corolla lobe bearing a protuberance. —F. Two adjacent flowers, showing
Breonia s
Figure 14.
Mature infructescence. —D. \
two adjacent ovaries separated: median
D. Calyx.
dissection through flower, showing a calyx and corolla tubes, and ovary (left);
—C.
Volume 89, Number 1
2002
Razafimandimbison
Revision of Breonia
species was identified by Capuron (1973b) as Breon-
ia chinensis (Lam.) Capuron; however, it is distin-
guished from B. chinensis by having bark with an-
nular fissures and by a single ovule per locule.
Paratypes. MADAGASCAR. Fianarantsoa: Dist
pura em 4855 SF (TEF); Forét de (ioe
23641 SF Toliara: District Fort-Dauphin, Forét
d'Analalava, cds 1367 (К, MO); Forêt d'llandy au Nord
de Fort-Dauphin, 20558 SF (TEF), McPherson 14391 (K,
MO, TEF); Manantenina, Beharena, 10925 SF (TEF).
19. Breonia tayloriana Razafim., sp. nov.
Madagascar. Toamasina: Fénérive Est,
Tampelo 3 Jan. 1956 (fl), 15703 SF (holotype,
TEF). Figure 15.
зесіеѕ a congeneris stipulis oe
арз ашаа carinatis perfacile distinguitu
Medium-sized trees, 9-10 m tall. Bark longitu-
dinally fissured. Leafy stems quadrangular, lenticel-
late, glabrous. Terminal vegetative buds conical, 13—
30 X 7-10 mm, glabrous. Leaves persistent, petioles
22-35 X 24.5 mm, adaxially canaliculate, gla-
brous; blades 18-32 х 10-15.5 cm, broadly obovate
to broadly ovate, reddish, glabrous, coriaceous,
glossy, apex obtuse to rounded, base cuneate; mar-
gins glabrous, entire; secondary veins 7 to 9 pairs
per side, eucamptodromous, prominulous; without
domatia; stipules 19-31 X mm, cymbiform,
united at the base,
semi-persistent. Inflorescences solitary (2 per axil),
heads 3.2-3.5 cm wide; inflorescence axes 3.24.2
cm long, strongly flattened, glabrous; bracts calyptra-
like, deciduous; peduncles 1—2 mm long. Flowers 5-
merous; calyx tubes ca. 2 mm long, green, inside
lanate, outside glabrous, lobes ca. 1.7 mm long, ob-
long to truncate, lanate; corolla tubes 8-9 X ca. 1
mm, light beige to cream, inside puberulous, outside
glabrous, lobes ca.
abaxially carinate, glabrous,
.8 mm long, oblong, at the lower
parts glabrous, puberulous toward the apex; anthers
ca. 1 mm long; filaments ca. 0.1 mm long, glabrous.
terete; styles ca. 13 X 0.2 mm, glabrous: stigmas
clavate; ovary 2-carpellate; carpels syncarpous:
ovules 2 to 4 per locule, strongly flattened, pendu-
lous, imbricate; placentae elongated. Infructescences
2-2.4 cm diam.,
nants; individual fruits with endocarp soft, fibrous;
with well-developed calyx rem-
disks accrescent, obconical; seeds 1 to 2 per locule
strongly flattened, ellipsoid, red; seed-coat reticulate.
Habitat and distribution. Evergreen rainforests
and littoral forests; Districts of Maroantsetra, Brick-
aville, and Fénérive Est (Fig.
Common names. Molompangady mena (red lips
of spade).
Phenology. Flowering December to February:
fruiting February to April.
Discussion. The specific epithet honors Char-
lotte Taylor, a Rubiaceae specialist who has studied
mostly the Neotropical Rubiaceae. Charlotte has
taught me various aspects of taxonomy, including
nomenclature, describing new species, writing a
monograph, herbarium curation, and specimen an-
notation. The present revision was carried out un-
der her supervision.
District
anton Ampasina, А Zavah Paul
TEF); District ae :'kaville, Ambila Lemaitso,
Ampanotoamaizina, 8317 SF (ТЕЕ); — jer nns "tra.
Masoala peninsula, coastal sand forest just N of Antalavia
(K, MO, P, TAN); Jardin жае de
Farankaraina, 5652 SF (TEF), 63- R-199 (TEF); [Locality
unknown], 32928 SF (TEF).
Paratypes. MADAGASCAR. Toamasina:
20. Breonia tsaratananensis hazafim., sp. nov.
ГҮРЕ: Madagascar. [Antsiranana province]:
Massif de Tsaratanana, haut bassin de la Bean-
drarezina (Andranomena), affluent rive gauche
de la Mahavavy, 2000-2300 m, 11 Nov. 1966
(fr), 27049 SF (holotype, TEF). Figure 16.
Haec species a congeneris tubo calycino supra medium
dilatato apicem basemque constricto facile distinguitur.
Trees, 15-30 m tall. Leafy stems quadrangular,
glabrous. Bark rugose. Terminal vegetative buds
conical, 8-9 X ca. 3 mm, glabrous. Leaves persis-
tent, petioles 15—20 mm long, adaxially canaliculate,
glabrous; blades 9.5-11.3 х 5.2-6.5 cm. elliptic to
obovate, brittle when dry,
glossy, apex mucronulate, base attenuate; margins
glabrous,
glabrous, coriaceous,
revolute, entire; secondary veins ca. 7
pairs per side, orange to yellow-tinged, eucampto-
dromous, prominulous; cryptic-type domatia at the
base of the secondary veins, evident, ovate, glabrous:
stipules ca. 10 X 3 mm, cymbiform, not carinate.
glabrous, deciduous. Inflorescences solitary, heads
ca 1.7 em wide; inflorescence axes 2.5-3.3 cm long,
slightly flattened, woody, densely pubescent: bracts
1-2
long, densely pubescent. Flowers 5-merous or rarely
calyptra-like, deciduous; peduncles ca. mm
ft
entire flower (right). —G. bu adjacent 2 — two adjacent carpels separated: median dissection of mature
infructescence, showin escent disk and s
lahimena 237 (MO) and G ae 11407 SF (TEF).
Seed, dorsal view (left); lateral view (right). A—F from Anti-
34 Annals of the
Missouri Botanical Garden
nt
кү AV
/ Vy
\ }
UN Jl
ur
M DM
Ñ {л ya
ГІ
May
Lu
V Eun En
NA ly ШУ
Ч ҮЛ (
anrs Ee —
mm
Figure 15. Breonia tayloriana.—A. Fertile branch with inflorescence and semi-persistent stipules. —B.
Two ad-
jacent flowers, showing tw
о adjacent ovaries separated: median dissection through flower, showing lanate calyx and
ovaries (left); entire flower (right). —C. Median dissection through a separate fruit, showing accrescent disk and ovaries
with young seeds. A-C from 15703 SF (TEF).
Volume 89, Number 1 Razafimandimbison 35
2002 Revision of Breonia
Figure 16. Breonia tsaratananensis.—A. Fertile branch with a young inflorescence and a mature infructescence.
—B. Two adjacent flowers, showing two adjacent ovaries separated: median dissection of calyx with adaxial view (left):
abaxial view (right). —C. Median dissection through fruit, showing accrescent disk and carpels with seeds (left); entire
fruit (right). —D. Dissected unilocular ovary. —E. Mature seed: dorsal (left); lateral (right). —F. Mature infructescence.
A, B from 973 SF (TEF) and C-F from 27049 SF (ТЕЕ).
Annals of the
Missouri Botanical Garden
4-merous; calyx tubes ca. 2.5 mm long, dilated
above the middle, constricted at both ends, promi-
nently ribbed, inside glabrous, outside puberulous,
lobes truncate, ribbed, quadrangular, tomentose; co-
rolla tubes 5-6 mm long, inside pubescent, outside
puberulous; lobes ca. 2 mm long, recurved, puber-
ulous; anthers not seen; style 7-8 mm long, glabrous:
stigmas clavate; ovary 2-carpellate, rarely 1-carpel-
late; carpels syncarpous; ovules 3 to 5 per locule,
flattened,
elongated. Infructescences 2-2.5 cm diam., with ac-
pendulous, imbricate; placentae large.
crescent calyx remnants, persistent on the branches
until the next flowering season; individual fruits with
endocarp soft, fibrous; disks accrescent, triangular,
seeds 3 to 4 per locule, concavo-convex, ellipsoid,
dark red; seed-coat reticulate.
Mid- and high-alti-
District Ambanja (including
Habitat and distribution.
tude humid forests;
Tsaratanana Mountain) (Fig. 12).
Common name. Valotro.
Phenology. Flowering December to February:
fruiting March to November.
This species is known from five col-
lections made from three individual trees; therefore,
it is considered to be rare. Also, this is one of the
two species of Breonia found only in high-altitude
(above 1500 m) humid forests.
The specific epithet refers to the highest Mala-
Tsaratanana (2880 m), where the
type species was collectec .
gasy mountain,
aratypes. MADAGASCAR. Antsiranana: District
Ambanja, Massif de ar vallée de la Mahavavy.
Morat 2306 (MO, TAN). hajanga: Mangindrano, Am-
bohimirahavavy, 97.3 SF (Ko MO, TEF).
DUBIOUS SPECIES
Breonia longipetiolata Havil.. J. Linn. Soc. Bot. 33:
36. 1897
definitely belonging to Breonia,
TYPE: Leprieur s.n. (P!). This specimen,
was reported to
have been collected by Leprieur from French Gui-
ana. However, the locality must be considered
doubtful because Breonia is restricted to Madagas-
car and has never been reported from South Amer-
-
^
Also, Leprieur never collected plants in Mad-
The of B.
longipetiolata does not match any of the species
agascar. type and only specimen
recognized here. This species is distinguished by
its leaf blades with acute to rounded apices and
long (ca. 9.6 em) and strongly flattened peduncles.
EXCLUDED SPECIES
Setch. =
(Setch.) Setch. & Christoph. This species was orig-
Breonia mayortt Sarcopygme тауоги
inally placed in Breonia simply because of its mul-
tiple fruits. It has been transferred to the genus
Sarcopygme of the tribe Morindeae based on the
presence of raphides and bifid stigmas.
Literature Cited
Abbiw, D. K. 1985. The site struc ture of Nauclea кж
ji eti . 497—498 in P. Goldblat
P. P. Lowry H (editors). um Systematic ga
e Апо an Botany. Monogr. Syst. I Missouri Bot.
Ww
=
=
rd, 2
— ” 1879. Observations sur les Nauclées. Adan-
sonia i 311-312.
. 1885. Liste des Se de M
ull. Soc. Linn. Paris 60: 17
Bakhuizen van den Brink, К. € 1970. Nomenclature and
— of the genera of йн 'eae-Naucleeae and
'oposal to conserve the generic name Nauclea L.
Taxon 19: 4608—44
d . R. 1995. Тһе Irporisnoe of the Black Le-
nur o m macaco, Lemuridae—Primates), for Seed
Dispecs al in Lokobe Forest. Madagascar. Doctoral The-
sis, University College, London, England.
Boiteau, P. 1985. Dictionnaire de noms Malgaches de vé-
ве taux, Microédition no. 85 04 10, Archive des docu-
‚ Muséum а d — ire Natur ене; Гап.
9
—
^^
Madagascar (suite).
>
ат. Neclamafokia, un nouveau nom pour ени
phalus auct. non A. Rich. (Rubiaceae). Bull. Mus. Natl.
ist. Nat., Sect. В, Adansonia, sér. 4, 6: 243-248.
1999, A propos d'Anthocephalus A. Rich. et de
Cephalanthus chinensis Lam. (Rubiaceae). Adansonia,
sér. 3, 21: 93-95.
т аи C. E. B. 1966. Remarks on the position, the
poe d and the subdivision of the Rubiaceae. Acta
. 15: 1
A. De. [Sep. E 1830. Prodromus Systematis Na-
Iv: Paris.
. Sur l'identité du ne chi-
аА sér. 2, 13:
73b. Revision des Rubiacées de Madagase ar
Z
a
nensis pu 71—47:
et dis м omores. Unpublished manuscript: Notes regrou-
I dacty — s
de F. Chauvet. Laboratoire Phanérogamie, P
1997. Plant Collectors in Mac — ar ane the
Co omoro Islands. ie Botanic Gardens, Kew.
Endlicher, S. F. L. . Enchiridion bu ‘um. Lipsiae
Sumptibus ( Gull. tim — Viennae.
Grandidiet y 1897. Histoire physique, — el
Маанаке аг Vol. 36: plate 457. Publié
Г Imprimerie Nationale, Pa
Guého, J. 1976. Sur l'identité de Phylica — ées)
des Masc are Ignes. — sér. 2, 15: 509-513.
Harris, J. G. € . 1994. Plant ш 'ation
Terminology. Bine la Ut tah.
Haviland, G. D. mee A revision of the tribe Naucleeae.
et mises en forme par J. Bos ser,
itique de
. Linn. Soc. . : 1-94
Heine, H. 1968. A propos de la nomenclature e ч Sé-
: — de l'ancien monde. Adansonia, sér. 2, 8: 181—
А А. М. 1938. Le genre Breonia de Madagascar.
Bull. Soc. 3o Fr, 84: 457—402
ds M.
s. Proc.
. On үш = newpoints — some
e Ned. Akad. W ser. С, 69: 275-316.
Volume 89, Number 1
2002
Razafimandimbison
Revision of Breonia
37
Lamarck, A. 1785. a Méthodique Bota-
sos Panc — ke, Paris.
789. о е Méthodique Botanique 4.
GA Paris. 435.
q E. Turner.
redator and A mites in leaf domatia. Amer.
J. Bot. 76: 105-112.
Radford, A. E., W. C. Dickison, J. R. Massey & R. Bell.
^a Vascular Plant Systematics. Harper & Row, New
Yor
байый. 5. G. & В. Bremer. 2002. Tribal de-
limitation of Naucleeae (Cinchonoideae-Rubiaceae):
Inference from melee ub and torba icd data. Syst.
Geogr. Pl. (in
Richard, A. [Dec. ] 1830. €— ү Rubiac: ‘Сев.
270 p.. reimpr. In Mém. Soc. . Paris 5: 81-304
1834).
(
Ridsdale, C. 1975. A synopsis of the African and Mada-
gascan د a ai eeae. Blumea 22: 541—553
1978. A revision of the tribe Naucleeae s. s. (Ru-
biac — 24: 307-3
8. Rubiaceae. Pp. п М. D. Dassan-
ayke hs — A Revised Тт... to de Flora of Cey-
lon, Vol. XIL A. A. Balkema, Rotterdam, Netherlands.
Robbrecht, 5 ee — woody Rubiaceae. Opera
Belg. 1 Г}.
A 1994. $ Saleen u to S 1988 outline of the clas-
sification of the Rubia . Index genera. Opera Bot.
Belg. 3(1993): 173- 196.
Schumann, К. 1891. Rubiaceae. In A. Engler & K. Prantl,
: 55-00.
893. Anatomischen C hora и und zur
Не. Boissier 1:277
éd.
“Mémoire sur la
famille des Rubiacées.” Pos O: 186 187.
Verdcourt, B. 1958. Remarks on the cla
ssification of the
l.
Rubiaceae. Bull. Jard. Bot. Bruxelles 28: 209-28
APPENDIX 1.
T OF SPECIES
. Breonia boivinii Hav
1 Breonia capuronii Razafin m.
3. Breonia chinensis (Lam.) Capuron
: Breonia cuspidata (Baker) Havil.
. Breonia decaryana Homolle
6 Breonia fragifera С ;apuron ех ее
‹
3
©
=;
&
8
=.
8
E
FP
&
3
&
a»
7
16. Breonia ik (Baill.) оке ex Ridsdale
17. Breonia stipulata Havi
18. Breonia taolagnaroensis Rasen:
19. Вгеота tayloriana Razafim.
. Breonia tsaratananensis Razafim.
INDEX TO EXSICCATAE
"xamined specimens are listed alphabetically by col-
lector, followed by collection numbers; the species is in-
dicated by a number in par
tenthe ‘ses corresponding to the
number in the List of Species
1989. Occurrence of
Andrianarisata et al. 259 (6); Antilahimena 19 (1),
. 237 (15); Antilahimena et al. 324 (6).
n 5563 (4), 6602 (4); Behasy & Vasey 203 (15);
Bernard 187 (3), 327 (3); Bernardi 11899 (1: 3); Birkin-
shaw 59(15), 67 (1); Barin s.n. (1).
P 3). s.n. (11); Commerson s.n. (3), s.n.
n. (18).
Decary 5199 (5), 5511 (3), 551
1367 —
Gautier 3772 (2); Gautier et al. 3288 (6).
Hildebrandt er (е; Humblot s.n. (3).
Jongkind et al. 3 13).
Labat et al. 2761 id 2263 (17). Labat & Conté 2674
(13); Lewis et al. 755 (9); Louvel 125 (8), E (12), 216 (5).
Malcomber 1219 (2), Malcomber et al. 1833 (13); Mc-
Pherson 14391 (18); Miller 3609 (17); Mora 2306 (20).
Noyes et al. 961 (6), 974 (10), 1028 (13).
12); Perrier de la Bâthie 351: 3 (13), 3904
1). 3933 (10). 5270 SF (10); eps s.n. (17).
Rabe 115 (15); Rahajasoa et al. 747 (3); posse : (16);
Rakotozafy & Raharilala 2189 (17), 2271 (17); Randri-
Hey" 241 (1), 251 (6), 255 (16); pape capi:
; Randrianasolo & Rasabotsy 32 (7 ); Ra
3 (3). 5573 (3); Dumetz
—
velonarivo
13 (13), SG 352 (10), SG 358 (7). SG 389 (8); Ré-
serves Naturelles: 2751 RN (1), 4905 RN (1), 5152 RN
‚ 5524 RN (15), 6234 RN (1), 6257 RN (16), 6809 RN
‚ 7822 RN (16), 7906 RN (2), 9002 RN (2), 9045 RN
(6), 9052 RN (10), 9129 RN (10), 9229 RN (16), 9453
RN (16), 9982 RN (3), 11089 RN (17), 12975 RN (2).
Schatz et al. 1912 Ue 3858 (3); Scott Elliott 2214 (4);
Station Forestiere: 841 SF (15), 973 SF (20), 1432 SF (1).
2078 SF (9), 2410 SF D 2950 SF (16), 2961 SF (6), 3026
E (13). 3265 SF (10), 3882 SF (13), 4677 SF (13), 4855
SF (18), 5652 SF (19), E. SF (3). 6105 y (7), 6181 SF
(13). 6360 SF (3), 6616 SF (13), 6696 У 13). — SF
(13), 6791 SF (17), 7136 SF (7), 7
7562 SF (8), 7921 SF (8), 7987 SF (1: H, 8245 p (13).
8317 SF (19), 8413 SF (15), 8734 SF (6), 9047 SF (3).
озы (15), 9238 SF (3), 9711 SF (9), 9974 SF (7), 10803
К (6), 10925 SF (18), 11407 SF (15), 11417 SF (16)
Ша SF (15), 11503 SF (6), 11594 SF (9), 12023 SF (17),
12352 SF (3), 12447 SF (3), 13094 SF (12). 14999 SF (14),
14437 SF (3), 14712 SF (3), 14728 SF (3),
15057 SF (17), 15273 SF (16), 15287 SF ( (3),
(3), 15495 SF (3), 15518 SF (3), 15673 SF (13), 15947 SF
‚ 16174 кый eb 16217 SF (3),
§ F (3), 17565 SF (13),
18116 SF ty les SF (14), 18450 SF (13), 18518 SF
SF (13), 19295 SF (7). 19379 SF
5 SF Ay 19791 SF (6), 20558 SF (18), :
SF ( Т), 26103 SF | 3), 26804 SF (5 E н
SF (20), 27192 SF (3), 27503 SF (13), 27593 SF
SF (1), 28081 SF (3), 28136 SF (10), 28414 SF (5). :
SF (5). 29653 SF (13). 31809 SF (3), 32928 SF (19). :
SF (10), 33237 SF (6), 33240 SF (3), 33536 SF (6). 3
SF (6). 34413 SF (6), 34711 SF (3), 34719 SF (3). S/N SF
(3).
Тюшген s.n. (о), 91 (5), 117 (10); Turk et al. 643 (3)
Villiers et al. 4846 (13); Vigreuse l 5418 (3).
y Е 524 I
avah 321-R-107 (19); 5524 RN (+
Unknown collectors: 12-R-B-172 (5): 14-B-R-230 (9);
86-R-118 (3); 63-R-199 (19); Herbier de la station agri-
cole de l'Alaotra, 2200 (TAN).
CLASSIFICATION, ORIGIN,
AND DIVERSIFICATION OF
THE NEW ZEALAND HEBES
(SCROPHULARIACEAE)'
Steven J. Wagstaff,? Michael J. Bayly,?
Philip J. Garnock-Jones,'
and Dirk C. Albach?
ABSTRACT
The New Zealand hebes үө rophulariaceae) are members of a large Southern Не — clade nested within Ve-
ronica. Analysis of ITS and rbcL sequences suggest s that
ancestor that arrived via long-distance dispersal.
Zealand, the hebes have unde "gone at
degree of morphological diversity in the New +
divergence. New Zealand w
high mountains or forest margins. Our oil sugges
was a source of ne
to South America,
~
supported by the sequence data.
Key w ords:
the N
After the establishment of this initial founder population in New
t least two major episodes of diversification,
| hebes contrasts with a
at least one instance from New Zealand to Australia, anc
ew Zealand species are derived from a single common
six clades. The great
агаш ыа low level of sequence
South Paci
giving rise to s
uinea have occurred relatively recently. Shorter hops to the Chatham Islands and the subantarctic islands are also
Hebe, ITS, New Zealand, phylogenetic analysis, rbcL, Serophulariaceae, Veronica.
Long-distance dispersal has a profound influence
on the evolution of insular floras (Carlquist, 1974),
and there is substantial evidence suggesting that it
occurs relatively frequently (Godley, 1967; Pole,
1994). One of the most remarkable examples of dis-
persal followed by adaptive evolution on islands is
the New Zealand hebes (Serophulariaceae). Wags-
taff and Garnock-Jones (1998, 2000) suggested that
the New Zealand hebes are the descendants of a
small founder population that may have been de-
rived from a single seed. They proposed that com-
bined influences of inbreeding, genetic drift, and
strong selection acting upon small populations have
probably played a major role in the rapid diversi-
fication of the group.
The hebes are one of the largest and most eco-
logically diverse plant groups in New Zealand, in-
cluding over 120 species, with outlier populations
in eastern Australia, Tasmania, New Guinea, Rapa
Island, and South America. They range from al-
pine cushion-forming plants (Fig. 5D) to lowland
woody shrubs or small trees (Fig. 5M, R, S), and
in New Zealand are conspicuous elements in most
terrestrial ecosystems except forests and wetlands.
Species such as Hebe armstrongii, H. cupressoides,
and H. speciosa have patchy or localized distri-
butions and are considered rare or endangered;
about 70% of the species are confined to small
regions within New Zealand.
The New Zealand hebes were formerly included
in a broadly defined circumscription of the genus
Veronica (Wettstein, 1891; Cheeseman, 1925), but
recent flora and taxonomic treatments (Ashwin &
1961; 1993a, b;
Heads, 1994a, b) recognize less inclusive groups
Moore in Allan, Garnock-Jones,
(see Table 1), usually accepting four genera in New
Hebe, Heliohebe, and Para-
1993a, b). Heads (1987) de-
scribed an additional genus, Leonohebe. Although
Zealand: Chionohebe,
hebe (Garnock-Jones,
we do not accept his wide circumscription of that
genus, the name Leonohebe could be applied to a
small clade of four or five species that is supported
by the analyses of Wagstaff and Garnock-Jones
(1998, 2000).
! The authors gratefully acknowledge curators of the experimental gardens and CHR ae TT
Research, Lincoln,
sequencing. Dick Olmstead contributed s
tics Ltd.) ا many of the photographs. a ind Alison
iefited gn — тот the ins
. This research was funde
along with the assistance of Anita Thorne, Andy Gla azie г, and Ё izabeth MacAvoy w
ightful comments of Christine Bezar,
the NA
Kellow assisted um preparation of figures. Earlie versions
Ilse Breitwieser, Mar k (
ew Zealand by the Foundation for Research, Science
thase,
d in Ne
and Tec :hnology (contrac : : 09618 dnd MNZOOI), 34 he Marsden Fund (е ontract LANOOI
2 Landcare Research, Box 69, Lincoln 8152
! Museum of New — Te Papa Tongarewa “PO. Box
v Zeal:
ind.
107.
.CrI. nz.
New Zealand.
арза
Welling glon,
! School of Biological Sciences 2d Island Biolog Research Programme: Victoria University of Wellington, P.O. Box
600, Wellington, New Zealand.
` Botanisches Institut der Universität Wien, Rennweg 14,
ANN. Missouni Bor.
GARD. 89: 38-63. 2002.
1030 Wien,
Austria.
Volume 89, Number 1 Wagstaff et al.
Diversification of the New Zealand Hebes
el. Classifications of New Zealand hebes. The New Zealand species were placed in three sections of Wettstein’s
akon treatment of Veronica, and Cheeseman (1925) placed them in two divisions of Veronica.
win & Moor
Ashw e
Wettstein (1891) Cheeseman (1925) in Allan (1961)! Heads (1987, 1994b) Garnock-Jones (1993a, b)
Veronica Veronica
sect. Pygmea Division Pygmea Pygmea Chionohebe Chionohebe Inc luding
Parahebe “Group B”)
sect. Hebe Le
onohebe
sect. Densifoliae
Division Hebe
Leonohebe
Connatae
ebe Heb
“Semiflagriformes” sect. “Semiflagriformes”
“Connatae” sect. “Connatae”
S
Ф
A
sect.
Apiti
“Flagriformes” Flagriformes “Flagriformes”
sect. Aromaticae
Salicornioides
hee n
sect.
“Buxifoliatae” sect. “Buxifoliatae”
“Subdistichae” sect. Sind hae ^| sect. Subdistichae
“Subcarnosae” sect. Glaucae sect. Glaucae
Heb
sect. Hebe sect
pertae” ser. Hebe ser. Hebe
“Occlusae” ser. Occlusae ser. Occlusae
Parahebe
“Grandiflorae” “Grandiflorae”
“Paniculatae” sect. Paniculatae Heliohebe
sect. Chamaedrys Division Euveronica Parahebe arahebe
Group A, B, C” “Group A, C”
sect. Labiatoides — — Derwentia
sect. Paederota =P == — —
sect. Paederotoides — — — —
sect. Pseudolysimachia — — — —
sect. Veronicastrum — — — —
sect. Omphalospora — => — =
sect. Beccabunga — — = =
! [n the Flora of New Zealand. — l erir 1961),
"mes
and the informal grouping "Flag
was prepared by L. B. Moc
This research contributes to ongoing efforts to
create a phylogenetic classification of Scrophulari-
aceae. Olmstead and Reeves (1995) and Olmstead
et al. (2001) showed that the Scrophulariaceae, as
traditionally circumscribed, are not monophyletic.
They identify clades from a dismembered Scrophu-
lariaceae s.l. that could merit formal recognition. In
their studies Veronica was nested within a large
clade they called ue
prop. (Reveal et al., his large clade was
recognized by — i Reeves (1995) and in-
cludes part or all of Bentham's (1876) tribes Digi-
taleae, Antirrhineae, Cheloneae, and Gratioleae,
the small tribe Angelonieae, and the small families
b ard nom. cons.
Callitrichaceae, Globulariaceae (excluding Selagi-
naceae), Hippuridaceae, and Plantaginaceae.
The aim of this research is to identify well-sup-
M. B. Ashwin prepared the treatment of Parahebe, Pygmea,
` of Hebe. The remainder of the Hebe treatment, including the informal synopsis,
ported monophyletic groups among the New Zea-
land hebes, to improve their classification, infer
their origin, and explore underlying processes of
diversification. We propose that diversification in
the group reflects transoceanic dispersal and adap-
tive radiation. The hebes have successfully exploit-
ed a diversity of ecological niches that were prob-
ably created during the recent uplift and glaciation
of the mountains of New Zealand.
MATERIALS AND METHODS
Our sampling strategy capitalized on the unique
characteristics of rbcL and ITS sequences. The
plastid encoded gene rbcL has relatively few vari-
able sites, which allowed sequence comparisons
among distantly related outgroups, and placement
Annals of the
Missouri Botanical Garden
of the hebes within Serophulariaceae. It is also use-
ful in that a large number of published rbcL se-
quences are available for comparison (see Chase et
al., 1993; Küllersjó et al., 1998; Olmstead et al.,
2001, and references therein). Finally, Albert et al.
(1994) and Bremer and Gustafsson (1997) suggest-
ed that the gene rbcL approaches clock-like behav-
ior in its evolution, and hence the amount of se-
quence divergence could be used to estimate
divergence times. By comparison, the nuclear en-
coded ITS-region has many more variable sites than
rbcL,
to resolve е at lower taxonomic levels
which provides more informative characters
(Baldwin et al.,
STUDY GROUP
The rbcL study group consisted of 33 species
including 12 of the New Zealand hebes with at least
l representative from each of the currently recog-
nized genera, 5 species of Veronica, and 1 species
Nineteen rbcL
~
=a
Veronicastrum. sequences were
newly published herein along with 13 published
sequences of Antirrhinaceae from Olmstead et al.
(2001), and Nicotiana tabacum (Solanaceae) was
designated as the outgroup (Lin et al., 1986). Seven
sequences were considered redundant; even though
they were not identical, the resolution of missing
data could potentially make them identical. We
therefore excluded Derwentia derwentiana and D.
perfoliata, and the New Zealand accessions of Hebe
elliptica and H. salicifolia, from subsequent analy-
ses. Nineteen of the 37 species included in the rbcL
analysis were also included in the ITS survey.
The ITS study group included 78 sequences, 19
of which were newly published. Among these are
58 representatives of Chionohebe, Derwentia, He-
liohebe, and Parahebe, including conspecific acces-
sions of Chionohebe ciliolata and С. densifolia from
Australia and New Zealand, Parahebe lithophila
from Australia, and P. randewateri from New Guin-
ea. Thirty-five species of Hebe were also included,
and among these were: at least one representative
from each of Moore's (in Allan, 1961) informal
groups; H. formosa from Tasmania; H. benthamii
from the New Zealand subantarctic islands; H. bar-
keri, H. chathamica, and H. dieffenbachii from the
Chatham Islands (east of the main islands of New
Zealand); accessions of H. elliptica from both New
Zealand and the Falkland Islands; and H.
folia from both New Zealand and Chile. Pseudoly-
Veronicastrum, and Wulfenia
salici-
simachion, Veronica,
emerged as potential sister groups of the hebes in
the analysis of Hong (1984) and Albach and Chase
(2001); therefore a total of 19 species representing
these genera were included in our analysis. The
Asiatic species Veronicastrum sibiricum was desig-
nated as the outgroup for the analysis of ITS se-
quences.
Voucher specimens are listed in Appendix 1,
along with collection information, literature cita-
tions, and GenBank (<http://www.ncbi.nlm.nih.
gov>) accession numbers. The complete data sets
are available upon — from the first author, and
TreeBASE (<http://www.
The study acces-
they were deposited ir
S capu O
sion number is $623, and the matrix accession
numbers are M961 (rbcL) and M962 (ITS).
DNA EXTRACTION, AMPLIFICATION, AND
SEQUENCING
Total DNA was extracted from either fresh leaves
or leaf fragments dried with silica gel using a mod-
ification of the hot CTAB method of Doyle and
Doyle (1987). The cpDNA gene rbcL and the
nrDNA ITS-region [the 3’ end of the 185 rDNA
gene; internal transcribed spacer —1 (ITS-1): the
5.88 rDNA gene; internal transcribed spacer ~2
(ITS-2); and the 5’ end of the 285 rDNA gene] were
amplified by PCR. Primer sequences and our am-
plification and sequencing techniques follow Olm-
stead et al. (1992) for rbcL, and Wagstaff and Gar-
nock-Jones (1998) for the ITS-region.
primers and unincorporated nucleotides were re-
moved from the PCR products by spin column cen-
trifugation (QIAquick PCR purification kit, QIA-
GEN Inc.). The purified DNA samples were then
labeled with Big Dye terminators (PE Applied Bio-
systems, The Perkin-Elmer Corp.). Both the forward
and reverse DNA strands were sequenced by the
Waikato University DNA Sequencing Facility. Con-
tig editing and assembly was accomplished using
Excess
Sequencher version 3.0 (Gene Codes Corp.).
SEQUENCE ALIGNMENT
The sequence alignment for the ITS-region was
facilitated by ClustalX (Thompson et al., 1997). A
gap penalty setting of 75 and a gap extension pen-
айу of 6.6 were initially used to identify and po-
sition large gaps in the sequence data; then low-
scoring segments were realigned using a gap
penalty setting of 15 and a gap extension penalty
of 6.6 with the removing new gaps option turned
on. These settings opened and positioned small
gaps. The final alignment was inspected and minor
revisions were made manually.
Volume 89, Number 1
2002
Wagstaff et al.
Diversification of the New Zealand Hebes
DATA ANALYSIS
The phylogenetic analyses were accomplished
using PAUP* version 4.0d65 (Swofford, 1998). The
analyses were conducted using the PAUP* settings
random addition sequence with 100 replicates.
TBR and
steepest descent. The characters were all unordered
branch swapping, mulpars in effect.
and weighted equally, and gaps were treated as
missing data.
Support for the inferred clades is given by jack-
1996) and by the
number of synapomorphies for each group. Jack-
knife percentages (Farris et al.,
knife analysis investigates the structure, or phylo-
genetic signal, in a matrix without permutation, but
excludes an assigned fraction of characters. here
set to 30%. The jackknife searches were performed
with 1000 replications excluding uninformative
sites, maxtrees = 10 for each replication; the start-
ing trees were obtained by random addition with
one replication for each jackknife replication, TBR
branch-swapping, and mulpars in effect.
The relationship between sequence divergence
and time for the gene rbcL was discussed by Albert
et al. (1994) and Bremer and Gustafsson (1997) and
was calculated using the equation:
substitution rate = patristic distanc -e(Dp)/number
of nucleotides/inferred time since cladogenesis.
RESULTS
The rbcL sequences were 1402 nucleotides in
length (positions 27-1428 in tobacco). Among the
1103 sites
were invariant; 144 were parsimony-uninformative,
1402 sites included in the rbcL matrix,
and 155 characters were parsimony-informative.
4.7%
Most of the missing data fell in a region at the 5’
Missing data accounted for of the matrix.
end of rbcL, upstream from the conserved EcoRV
restriction site used in cloning some sequences
(Olmstead et al.,
from position 1325, the location of a PCR primer
1992). and in a region downstream
site used to amplify some sequences. Most changes
(calculated across the maximum parsimony tree
shown in Fig. 2) occurred in the third codon posi-
tion (369); substantially fewer changes occurred in
the first (116) and second (47) codon positions.
Changes in the third codon position are generally
synonymous, and hence are more likely to evolve
in a clock-like manner.
The analysis of rbcL sequences recovered 9408
maximum parsimony trees distributed in a single
island of 529 steps (consistency index = 0.52 ex-
cluding uninformative characters, retention. index
= 0.73); a
strict consensus tree is shown in Figure
l, and one of the maximum parsimony trees 15
shown in Figure 2. Most members of the Antirrhin-
aceae form a basal grade in our analysis with Chio-
nohebe, Derwentia, Hebe, Heliohebe, Parahebe, and
Veronica forming a clade that receives 100% jack-
knife support. V. anagallis-aquatica is sister to a
largely Australasian clade (98% jackknife support)
that includes the New Zealand hebes along with
Derwentia, Veronica arguta, and V. persica. Rela-
tionships within this clade are poorly resolved
Figs. l,
5
p
The mean absolute distance and standard devi-
ation from Veronicastrum sibiricum to the Chionoh-
ebe, Derwentia, Hebe, Heliohebe, Parahebe, and Ve-
19
EN
-
ronica terminals is (see Fig.
Veronicastrum is reported in the fossil record (Tiff-
ney, 1985) from the mid Miocene some 15 million
years before present (mybp). The substitution rate
in the Veronicastrum lineage was estimated by di-
viding 29.1 + 7.7/15 = + 0.5 substitutions for
the entire gene rbcL. The mean distance from the
terminals to the ancestral node of the Australasian
species (including V. persica, which is Eurasian) is
18.8 changes, which corresponds to an upper Mio-
cene divergence estimate of about 9.9 mybp. The
mean distance from the terminals to the ancestral
node of the H. salicifolia lineage is 7.4 changes.
which corresponds to a Pliocene divergence esti-
mate of about 3.9 mybp (Fig.
The aligned ITS matrix was 695 nucleotides in
length with gaps created to account for insertions
and deletions, among which 364 sites were con-
stant, 109 were parsimony-uninformative, and 222
were potentially parsimony-informative. Missing
data accounted for 1.2% and gaps accounted for
9.6% of the ITS data matrix. The 5.85 gene was
uniformly 165 nucleotides; ITS-1 varied between
175 and 224 nucleotides and ITS-2 between 203
and 215 nucleotides. Most of the p in the
ITS region was observed in ITS-1 and ITS-2. The
5.88 gene was more conserved. Chuserved Dur
identified by Liu and Shardl (1994) and Hershkov-
itz and Zimmer (1996) were identified in the ITS-
| and ITS-2 sequences in our survey.
Thirty-eight insertions and deletions (indels)
were inferred in the ITS-1 and ITS-2 spacer regions
—
Table 2). Mostly these were relatively small, rang-
ing from 1 to 3 bp, but two large insertions of 27
and 44 bp and three deletions of 7 or 8 bp were
also inferred. Most of the indels were unique to a
single sample, but 13 were shared by two or more
species, sometimes uniting groups supported by
substitutions in the sequence data alone. Species
of Derwentia are characterized by a one-base de-
letion. Species of Heliohebe are characterized by a
42 Annals of the
Missouri Botanical Garden
Hebe elliptica
= Hebe benthamii
Hebe odora
Hebe salicifolia
Hebe salicornioides
Parahebe vandewateri
Hebe macrantha
Hebe cupressoides
Heliohebe raoulii
Chionohebe densifolia
Hebe cheesemanii
New Zealand hebes
» Parahebe catarractae
68 Derwentia nivea
Veronica arguta
Hebe formosa
100 — Veronica catenata
Veronica persica
Veronica anagallis-aquatica
Veronica officinalis
Digitalis purpurea
79 Plantago lanceolata
Veronicastrum sibiricum
[ — Callitriche heterophylla
85 Hippuris vulgaris
98 [ — Chelone obliqua
Collinsia grandiflora
= Antirrhinum majus
Globularia cordifolia
100 Gratiola pilosa
Amphianthus pusillus
Bacopa caroliniana
Angelonia pubescens
Nicotiana tabacum
Figure |. Strict consensus of 9408 minimal — trees produc ed by parsimony analysis of rbcL sequences. This
tree shows the placement of New Zealand hebes within the A ae sensu Olmstead et al. (2001), using Nicotiana
tabacum as an outgroup. Jackknife values > 50% are given above each node.
Volume 89, Number 1 Wagstaff et al.
2002 Diversification of the New Zealand Hebes
Hebe benthamii
Hebe salicornioides
Hebe odora
Hebe elliptica
3.9 Hebe salicifolia
mybp Derwentia nivea
Veronica arguta
H Parahebe catarractae
Parahebe vandewateri
— Hebe cupressoides
[ Hebe macrantha
Chionohebe densifolia
Veronica catenata
Veronica persica
Hebe formosa
— Hebe cheesemanii
- 9.9 Heliohebe raoulii
mybp Veronica anagallis-aquatica
V
eronica officinalis
Veronicastrum sibiricum
"n Digitalis purpurea
Plantago lanceolata
Callitriche heterophylla
Hippuris vulgaris
Globularia cordifolia
Chelone obliqua
Collinsia grandiflora
Antirrhinum majus
[ — — Gratiola pilosa
l
Amphianthus pusillus
Bacopa caroliniana
Angelonia pubescens
Nicotiana tabacum
— 5 changes
иге 2. One of the maximum parsimony trees recovered from a —— Кар of rbcL sequences. The branch
lengths are proportional to the number of changes along eac E — 1. See scale at bottom. Fossils of Veronicastrum
sibiricum are reported from the mid Miocene about 15 m bê. ence RE are xq at the ancestral node
TB
of Australasian species (V. persica is Eurasian) and the anc sl udi of the Hebe clade
Annals of the
Missouri Botanical Garden
€
Table 2
Insertions and deletions inferred from ITS
sequence comparison.
Insertion/
Taxon etion Size Position Sequence
Veronicastrum sibiricum Deletion 1 41 Uc
Veronicastrum sibiricum, Wulfenia Insertion | 54 с
carinthiaca
Veronica glandulosa Deletion | 67 Ug
Hebe vernicosa Insertion 3 68-79 gla
Veronica calycina Insertion | 77 t
Veronica chamaedrys Deletion l 78 c/g
Veronicastrum sibiricum Deletion 1 89
Wulfenia carinthiaca Insertion 44 90-134 aatctagglgtgcaagcccctttgttgagag-
lccegegectgetc
Pseudolysimachion spicata, Veronica Insertion 27 108-134 gaclagicgagtececcgcictegete
anc Moris ipd V. bellidioides, V
glandulosa, V. glauca, V. officinalis, \
serpyllifolia, V. urticifolia
Veronica bellidioides, V. officinalis, \ Insertion l 162 a
urticifolia
Derwentia derwentiana, D. nivea, D Insertion l 207 a
perfoliata, Parahe * ое.
Veronica arguta
Veronica macrostachya Deletion 3 208-209 ce
Hebe salicifolia Insertion l 17 c
Veronica bellidioides, V. glandulosa, Insertion l 234 c/g
V. officinalis, V. urticifolia, Wulfenia
carinthiaca
Veronica persica Deletion l 248 a
— ed ‚ anagallis- Insertion l 260 a
aque
P. — hion spicata, Veronica Insertion l 442 a
chamaedrys, V. macrostachya, V.
oltensis
Veronica urticifolia Insertion l 454 с
Veronica anagallis-aquatica Insertion l 467 t
Pseudolysimachion spicata, Veroni icc Deletion l 474 g
anagallis-aquatica, V. belli divides,
fruticulosa, V. glandulosa, y ow
V. officinalis, V. saturejoides
serpyllifolia, V. urticifolia,
Veronicastrum sibiricum, Wulfenia
carinthiaca
Parahebe canescens Insertion 1 47: g
Veronica bellidioides Insertion 2 501 сс
ulfenia carinthiaca Deletion l 503 |
Heliohebe hulkeana, Н. lavaudiana, Н. Insertion l 508 [
raoulii
Veronica persica Deletion l 513 t
pendens H. elliptica, Parahebe Deletion 8 583-590 tetegtge
‚ P. brevistylis, P. decora, P. lyallii,
rathulata, p vandewateri
Parahebe planopetiolata Deletion 7 590—596 саісісе
Parahebe canescens Deletion 8 591—598 atctecge
Veronica persica Deletion 3 592—594
Heliohebe hulkeana, H. lavaudiana, H. Deletion l 606 g
raoulii
Parahebe vandewateri Deletion l 609 a
Veronica chamaedrys Deletion 2 609-610
Volume 89, Number 1
2002
Wagstaff et al. 45
Diversification of the New Zealand Hebes
Table 2. Continued.
Insertion/
Taxon deletion Size Position Sequence
Parahebe canescens Insertic 3 615—616 cat
Veronica chamaedrys Deletion 2 620-621 te
Pseudolysimachion spicata Jeletion 2 625-626 ac
Veronica chamaedrys, V. Insertion ] 631 a
Veronica austriaca, V. olter Insertion 2 638—639
Veronica arguta Insertion 1 640 c
one-base insertion and a one-base deletion. Both
accessions of Hebe elliptica and nine species of
Parahebe have an eight-base deletion that appears
to have evolved independently at least three times
(Table 2). This deletion is lacking in Parahebe lin-
ifolia and in P. catarractae subsp. catarractae and
subspecies martinii. The South American accession
of Hebe salicifolia has a unique one-base insertion
that is lacking in the accession of Hebe salicifolia
from New Zealand.
The ITS
a faster rate than rbcL. The average rate of change
per variable site for rbcL was 1.7 (tree length of
532/number of variable sites 144 + 155). The
aligned ITS sequences were shorter than rbcL,
sequences in our study were evolving at
there were more variable sites (331), and the av-
erage rate of change per variable site was 3.6 for
the ITS region.
Parsimony analysis of the ITS-region recovered
6931 maximum parsimony trees distributed in at
least two islands of 1213 steps (consistency index
— 0.41 excluding uninformative characters, reten-
tion index = 0.73); a strict consensus tree is shown
in Figure 3 and one of the maximum parsimony
trees in Figure 4. Based upon the results from anal-
ysis of rbcL sequences (Figs. 1, 2). Veronicastrum
sibiricum was designated as the outgroup. The ear-
liest divergence within the ingroup is between Wul-
fenia carinthiaca and all other taxa. The Northern
Hemisphere species of Veronica are found in five
clades that form a grade basal to a Southern Hemi-
sphere clade comprising the New Zealand hebes
and their relatives. A heterogeneous Australian
clade comprised of Derwentia, Hebe formosa, Par-
ahebe lithophila, Veronica arguta, and V. calycina
(8296 jackknife; 18 synapomorphies) is sister to the
New Zealand hebes, though there is relatively little
support for this relationship (70% jackknife; 10
synapomorphies) (Figs. 3.
Six well-supported clades are identified among
the New Zealand hebes, but the relationships
among these clades are unclear (Figs. 3. 4). The
first is a clade that comprises Leonohebe s. str. in-
cluding a well-supported group. Hebe tetrasticha, H.
cheesemanii, and Н. ciliolata (99% jackknife: 6
synapomorphies) with Hebe cupressoides weakly
supported as their sister (61% jackknife; З syna-
he Chionohebe A clade (98% jack-
knife; 5 synapomorphies) includes Parahebe plan-
pomorphies).
opetiolata and the cushion-forming species of
Chionohebe with both the New Zealand and Aus-
tralian accessions of C. ciliolata. The Chionohebe
B clade consists of Parahebe trifida and both the
New Zealand and Australian accessions of Chio-
nohebe densifolia (91% jackknife: 4
phies). The fourth clade includes 6 species of Par-
ahebe (93% jackknife;
accommodates the informal “Groups
Ashwin (in Allan, 1961; Table 1) and P. spathulata.
The fifth clade includes all the species of Heliohebe
in our analysis (100% jackknife; 13 synapomor-
synapomor-
7 synapomorphies) and
& C" of
phies). The sixth includes the remaining species of
Hebe with both New Zealand and South American
accessions of Hebe salicifolia and Hebe elliptica
(100% jackknife; 11 synapomorphies). Hebe ma-
crantha is weakly supported as the sister to the rest
of this clade (50% jackknife; 4 synapomorphies)
(Figs. 3, 4).
DISCUSSION
arge, unwieldy genera with a cosmopolitan dis-
tribution such as Veronica pose among the most dif-
ficult taxonomic problems for plant systematists,
whose opinions are often strongly held. One of the
most vexing of these problems is the inconsistent
means by which taxonomists define generic bound-
aries and the recognition of rank within a hierar-
chical classification scheme. Recent taxonomic
treatments in the Southern Hemisphere have fa-
vored narrow circumscriptions, and several new
genera have been segregated from Veronica (see Ta-
ble 1), whereas taxonomists in Europe and North
America have traditionally embraced a broad ge-
46
Annals of the
Missouri Botanical Garden
ebe elliptica
10 = Дере die bachi
ebe chüthamica
| ep € ockayneana
I Pee Salic gonis
Hehe pimeleoiqes
townsonii
ebe odora
Де е pauciramosa
ebe epacridea
ере ramosissima
ebe elliptica
Hebe
arkert
2 e saticornioides
ae armstrongi
Heliohebe
Parahebe
VG
cheesemanii
be
2 be densifolia
] Chionohebe B
di MA al 2 Es i АШ
irane ebe ir
Chionohebe A
Leonohebe
E]
|
aden е tana
perfoliata
je PW e 20d Ауга
'arahebe oe
Hebe e formosa
9 eronica austri
86 Veronica o a mac rostac hya
d * ^
100
а алак», aids a
Veronica serpyi —
о та carinthiaca
onicastrum sibiricum
Stric
Figure 3.
Notable clades are ‚ identifie d with brackets. Jackk
nife v
ractae subsp. martinii,
subsp. hectoril.
Wettstein (1891).
This discrepancy of opinion contributes to taxo-
neric definition of Veronica, e.g.,
nomic ambiguity and instability.
Our results support those of Albach and Chase
(2001). implying that the genus Veronica is at best
paraphyletic by exclusion of the Southern Hemi-
sphere genera Chionohebe, Derwentia, Hebe, He-
liohebe, Leonohebe, and Parahebe (Figs. 1. 3). as
'onsensus of 6931 minimal length trees produc ed by parsimony analysis of the entire ITS
values >
“Parahebe catarractae subsp. rias: ‘Hebe hectorii subsp. subsimilis, ‘Hebe hectorii
-region.
90% are given above each node. 'Parahebe catar-
well as the Eurasian genera Paederota and Pseu-
dolysimachion, and the North American genera
Synthyris and Besseya. One possible solution is to
lump them all in a broad circumscription of Ve-
ronica. This move, however, would create a cas-
cade of nomenclatural changes requiring the rec-
ognition of many new combinations and the
adoption of old combinations within Veronica. We
Volume 89, Number 1 Wagstaff et al.
2002 Diversification of the New Zealand Hebes
Hebe .
Mieres Chile |
be pimeleoides
ё fownsonu
€ ana
Uo ера i yin
ы : en
ы! е elliptica slands
Бес dt лас
E Heb be phonic Chatham
v propinqua Islands
Hebe сорда 4
eb opodigides
he
п И
be ar ib on i Su bantarctic
e * niha Islands
rantha
ma
Lei ШОТ
y 1 zhe 'aoulii
ebe
pe hookeriana 4
Parahebe c atarractae
rc
arahebe canescens
nee ln li nifolia |.
р ara
New Zealand hebes
AAA ter Я
Parahebe "che seman a New Guinea
f
Chie — н Ар Australia
trifida
инн) e tham
uon 5h sbe thomsonii
Chionohe be pulvinaris .
С Tuonohefe ciliolata Australia
Chionohebe ciliolata
Par ahebe planopetiolata
(13 tetrasticha
ciliolata
‘heesemanii
ebe с “upressoides a
Ver onica arguta
Derwentia derwentiana А
De De. лла perfoliata Austral 1a
nivea
Parahebe lithophila
Hebe (pi
^ronica austriaca =
eronic a mac Tasia hya
Veronic
гон а ре r SIC a
ronica chamaedrys
[
Veronica fruticulosa
'eronica edo
стая мө spicata E "
onicq glau urasi
Ver — а be vate 1014 es asia
nica officinalis
ía
т а glandulosa
eronica anagallis- їїса
—— lo la
Wulfenia carinthiaca
Veronicastrum sibiricu
Ver
Ve ronjcan a ur бас
— 5changes
One of the maximum parsimony trees recovered from a parsimony analysis of the ITS- region. The New
nrbs KR comprise a well-supported monophyletic group with outliers on the offshore islands and in Australia,
New Guinea, and South America. Dispersal away from the main islands of New Zealand is inferred in nine species.
Branch lengths are proportional to the number of changes along each branch. See scale at bottom. 'Parahebe catarractae
subsp. martinii, *"Parahebe catarractae subsp. catarractae, *Hebe hectorii subsp. subsimilis, ‘Hebe hectorii subsp. hectorii.
sequences and describe patterns of diversification
accept that retaining a paraphyletic Veronica ob-
in the New Zealand hebes.
scures phylogenetic relationships: however, an al-
ternative approach is to recognize smaller, less in-
clusive clades as generic segregates of Veronica.
This approach was adopted by Hong (1984). Here A heterogeneous clade composed of Derwentia,
Hebe formosa, Parahebe lithophila, Veronica arguta.
MAJOR CLADES OF NEW ZEALAND HEBES
we identify major clades supported by the DNA
Annals of the
Missouri Botanical Garden
Figure 5. Plate illustrating some of the morphologic al diversity В * New Zealand hebes. А. Flowering shoot
of Derwentia pe rfoliata with toothed leaves obscure in this pict —B. Flowers of Parahebe catarractae. —C. Panic-
ulate and terminal inflorescence of Heliohebe raoulii subsp. — P. J. Garnock-Jones 2123. —D. Cushion habit
Volume 89, Number 1
2002
Wagstaff et a 49
|.
Diversification of the New Zealand Hebes
and V. calycina is sister to the New Zealand hebes
(Figs. 3, 4). The genus Derwentia includes nine
currently accepted species that are endemic to
Australia, from southeastern Queensland to Tas-
mania and west to Kangaroo Island in South Aus-
tralia, where they are found mostly in tableland or
cool temperate regions (Briggs & Ehrendorfer,
1992). Toothed leaves [only on the lower branches
of H. formosa and often obscure in D. perfoliata
(Fig. 5A)], and a dense ring of hairs in the corolla
throat (glabrous in H. formosa) are possible synap-
omorphies that unite the clade. Derwentia and Hebe
formosa also have similar growth forms; their new
shoots are initiated at the base of the plant, over-
topping older, short-lived branches. Hebe formosa
is distinguished from Derwentia by the occurrence
of a one-base insertion (Table 2
number of Hebe formosa is n = 21, this possibly
being the ancestral state within the Derwentia
clade; chromosome numbers of n = 19 or 20 are
published for Derwentia (Briggs & Ehrendorfer,
1992)
. The chromosome
We identify six major clades within the New Zea-
land hebes that are supported by the sequence
data, which we refer to as the Leonohebe clade, the
Chionohebe A clade, the Chionohebe B clade, the
Parahebe clade, the Heliohebe clade, and the Hebe
clade (Fig. 3). ITS sequences provide strong sup-
port for the Leonohebe clade (Figs. 3, 4), comprising
Hebe cheesemanii (Fig. 5K), H. ciliolata (Fig. 5J).
and H. tetrasticha. Hebe cupressoides (Vig. 5L) is
weakly supported as sister to this clade. These spe-
cies are endemic to the South Island of New Zea-
land. They have traditionally been included їп
Hebe, but here, as in previous analyses (Wagstaff &
Garnock-Jones, 1998, 2000), are far removed from
the other species that Moore (in Allan, 1961) in-
cluded in that genus. Hebe cupressoides and mem-
bers of the Leonohebe clade lack the dorsal capsule
compression typically found in members of Небе,
and a potential synapomorphy for the group is pos-
session of a distinctive cupressoid growth habit,
which has apparently evolved independently from
that in the whipcord hebes [e.g.. Н. annulata, Н.
armstrongii, H. hectorii, H. lycopodioides, H. pro-
pinqua, and H. salicornioides, which have a similar
growth form (Figs. 5T-W)]. Hebe cheesemanii, H.
ciliolata, and H. tetrasticha, along with H. tumida,
comprise the informal group “Semiflagriformes” «
Moore (in Allan, 1961). These similar species are
subshrubs that occur in rocky areas at high altitude
and are characterized by possession of lateral in-
florescences (Fig. 5K), dioecious breeding system.
and sour-scented flowers. We refer to this group as
E
the Leonohebe clade, because Leonohebe ciliolata
(H. ciliolata) was designated as the nomenclatural
type for that genus by Heads (1987). If the group.
including H. cupressoides, is treated as generically
distinct from Hebe (and other genera of the Hebe
complex; see Garnock-Jones, 1993a), then the use
of the name Leonohebe seems warranted, though
this is a much more restricted use of the name than
the clearly polyphyletic circumscription originally
employed by Heads (1987).
The Chionohebe A clade encompasses the cushion-
forming species of Chionohebe. All of these species
occur in the South Island where they are high-alpine
plants of rock and scree. The cushion-forming species
tft
of Chionohebe thomsonii, male plant, Eyre Mts.,
South Island,
P. J. Garnock- Jones "d —E. Lateral view of
erect tubular flower of че erus pulvinaris, from Takitimu Range, South Island. v а: var.
brachyphylla. —F. Habit of a plant ca. 15 em tall from same population as M. Bay » 560. rthur, South
Island. —G. Apex of v — shoot from Mt. Arthur, WELT 82554. H, I. He he фай п —Н. үч. of a plant
ca. 10 cm tall, same population as M. J. Bayly 795, Mt. St. Patrick, South Island. —I. Brane hlet of ү ae unknown
locality. —J. Hebe — branchlet showing ciliolate leaf margins, plant from Mt. Arthur, WEL ES ebe
a plant, from same аи as M. J. Bayly 756-757, Black Birch Ra., des Ed —L. Hebe
cupressoides, ui anchlet, bern ated , Nelson, South Island, WELT 82553. M, N. Hebe parviflora. —M. Habit of a
plant ca. 2.5
Р) J. — d 2257. O, P. F
/ vu bud, from Mt.
J. Bayly S-31.
speciosa, — from a plant cultivatec
S. Hebe salicifolia, shoot of plant from Upper Wairau
a
Hebe albicans. —O.
subsp. lycopodioides. —
qos portion of vegetative branchlet, cultivé
rndale, South Island, WELT 82551. Үү
2T. ‚ Lake Tennyson, South Island. —W.
vated i a Landcare Research Gardens, Lincoln, үм к
T lon Scale bars: A, B, m; C, G, I,
tall from same Со as Y J. Garnock-Jones 2258, Hauhangaroa Ra., North Islan
Shoot apex — young latera
Arthur, South Island, WELT 82555.
‚ Hebe O shoot of plant from М. | b South Island,
at Otari-Wiltons Bush, V
Habit of plant — same popula
d. —N. Inflorescence,
inflorescences and large apical
Shoot of plant from Cobb ү South Island, М.
M. J. B iyly 1478. —R. Hebe
Wellington, — — Maunganui Bluff, North Island.
У. Н
Valley, South Island, M. J. Bayly Hebe ү
ition as M. J. y O, Ms HEN, South Island. —U.
care ا Garde ns, Lincoln, originally from Fish Lake,
nce of a plant from same population as M. J. Bayly
)ranc na ч Ice a showing terminal infructescences, cu ы -
1 Nigger Stream, Canterbury, South Island, WELT 8255
mall, light seeds of Hebe elliptica (by pic 'al o PN of most Hase] cultivated, Otari-Wiltons Bush, ier n
N = 5 mm; E, V. Q, К, 5
= 2mm: J, К, U; X = I mm O, Р,
Annals of the
Missouri Botanical Garden
of Chionohebe (e.g., Fig. 5D) are united by several
synapomorphies, including cushion habit with the de-
cussate leaf pairs slightly offset to form a pseudo-
spiral (Heads, 1994b), thick-walled bristle-like eglan-
dular hairs, solitary flowers with erect. long corolla
tubes (e.g., Fig. SE), and corolla veins branching dis-
tally in the tube (Garnock-Jones, 1993a). The Chio-
nohebe A clade is sister to Parahebe planopetiolata,
“Group ol As
win (in Allan, 1961). With Parahebe planopetiolata
they share a five-lobed corolla and hygrochastic
—
>)
one of the species of Parahebe
capsule dehiscence, but the species in the Chio-
nohebe B clade also have these features.
The Chionohebe B clade includes С. densifolia
(Australian and New Zealand populations), plus
Parahebe trifida from the informal Parahebe
"Group B” of Ashwin (in Allan, 1961) (Figs. 3, 4).
Parahebe birleyi is not included in this clade in all
trees, and in the consensus tree forms a polytomy
with it and several other clades in the complex. The
species of the Chionohebe B clade are all alpine
plants found in the southern region of the South
Island. Parahebe trifida occurs in alpine flushes
and snowbanks, P birleyi is a plant of nival rock
ledges, and C. densifolia is found in a range of
stony alpine habitats. Chionohebe densifolia is also
found in the Kosciusco National Park in Australia.
Chionohebe densifolia, P. trifida, and P. birleyi are
all similar in appearance. Several of their shared
characters are likely to be synapomorphies, includ-
ing few-flowered inflorescences, large flowers with
purple anthers, and presence of long glandular
hairs on leaves. Other shared characters are also
shared with the Chionohebe A clade, including old
leaves withering and fading but retained on stems.
All three species of the Chionohebe B clade are
thought to form hybrids with cushion-forming spe-
Wagstaff & Garnock-
There is little evidence from mor-
cies of Chionohebe A clade (
2000).
phology to separate the Chionohebe A and B clades.
Garnock-Jones (1993a), from a cladistic analysis of
morphological and flavonoid data, proposed a more
Jones,
inclusive “Chionohebe” clade incorporating all the
species of Chionohebe and Parahebe “(
—
Group B” «
Ashwin (in Allan, 1961). Such a grouping appears
paraphyletic at best in this and earlier ITS studies
(Wagstaff & Garnock-Jones, 2000).
The Parahebe clade includes all the represen-
tatives of the informal Parahebe “Group A" of Ash-
win (in Allan, 1961) plus P. spathulata. Parahebe
spathulata is anomalous in this clade, which oth-
erwise has morphological support from inflores-
cence, floral, and flavonoid characters (Garnock-
Jones, 19934).
clade and P canescens, the sole species of Ashwin's
A sister relationship between this
"Group C." has 9396 jackknife ‘Support. The infor-
mal “Parahebe Groups A and C” share several flo-
ral apomorphies (Garnoc igne 1993a). Parahebe
brevistylis and P. linifolia (from Ashwin's “Group
В”) form a weakly supported small clade, which is
sister to the “Group A and С” The clade
is represented in both the North and South Islands
f New Zealand,
species.
^
©
where species occur in well-
drained soils associated with river banks, cliffs, and
screes. Parahebe canescens is a creeping diminutive
herb of South Island lake shores; its reduced fea-
tures match convergent similarities seen in other
plants associated with this habitat. The entire clade
except for P. brevistylis and P. spathulata is united
by floral features such as short corolla tubes, col-
ored nectar guides (Fig. 5B), and stamen filaments
narrowed at the base. In the case of P. brevistylis,
the differences can be explained as losses of ad-
aptations for insect pollination (Garnock-Jones,
1976b). Parahebe spathulata shares some features
of habit and flower morphology with P. cheesemanii
and might have an allopolyploid origin involving
species from the Parahebe clade and the P. chee-
semanii lineage.
ıe Heliohebe clade (Figs. 3. 4) was formerly
recognized as Hebe "Paniculatae" in the informal
in Allan, 1961), and later
segregated as a distinct genus by Garnock-Jones
classification of Moore
—
(1993b) (Table 1). It was also previously recognized
as a distinct group in the key of Cheeseman (1925).
Heliohebe includes five species that are found in
northeastern parts of the South Island on rock out-
crops, cliffs, and sometimes in grassland. Mono-
phyly is well supported by two unique indels and
several possible morphological apomorphies in-
cluding an inflorescence that is a terminal, com-
pound raceme or spike (Fig. 5С), protogyny (also
evident in Fig. 5C, where styles are protruding from
buds on the lowest inflorescence branches), sta-
mens erect, anthers cream or yellow, seeds fusiform
to irregular in shape and winged, and hemitropous
ovules (Garnock-Jones, 1993b).
The Hebe clade corresponds to Hebe sensu Moore
(in Allan, L961) with the exclusion of Heliohebe and
members of the Leonohebe clade. The majority of
species within the Hebe clade form a well-support-
ed group (100% jackknife value), with weaker sup-
port for Hebe macrantha (Fig. 5F, G) and H. petriei
as sisters to this group (Figs. 3, 4). The Hebe clade
is largely endemic to New Zealand, including many
of its surrounding islands, with two species also
extending to South America, and one species (H.
rapensis, not included in this analysis) endemic to
Rapa Island (Fig. 6). The clade is large and mor-
phologically diverse; unambiguous morphological
Volume 89, Number 1
2002
Wa
gstaff et al. 51
Diversification of the New Zealand Hebes
+ س Y
Mas Le 80° э
^
e PP a
| 2 Ри ae e^
е y
\
\
4 Kermadec
' Islands
MEE
i =
\
\ * Chatham
Islands
Figure 6.
synapomorphies are difficult to identify, with ab-
sences or reversals in some taxa. Wagstaff and Gar-
nock-Jones (1998) suggest that synapomorphies
may include: a shrubby or arborescent habit (e.g..
Fig. 5M, R, S), large leaf bud (e.g..
leaf margins, protandrous flowers, peltate placentas,
Fig. 50). entire
acute capsule apices, 3/5 or 5/8 inflorescence phyl-
lotaxis. Within the Hebe clade there is little reso-
lution, but several relationships are worthy of note.
Firstly, the “Connatae” of Moore [in Allan, 1961,
represented by H. benthamii, H. petriet, H. epacri-
dea, (Fig. 5H. 1) and H. ramosissima] are polyphy-
letic; with some members closely related to “Bux-
ifoliateae” (H. odora and Н. pauciramosa). These
relationships were first suggested by Wagstaff and
Garnock-Jones (1998),
taxa in this study continues to support this earlier
and the addition of further
assessment. Secondly, as found by Wagstaff and
Wardle (1999), three of the cupressoid species, Н,
salicornioides, Н. armstrongii (Fig. 5W), and H. an-
nulata (which share possession of fused anterior
21),
Thirdly, two of the
calyx lobes and chromosome number of n =
form a well-supported clade.
Chatham Island endemics, H. chathamica and H.
3, 4). with
dieffenbachit, are sister species (Figs.
Map showing postulated dispersal of hebes from the main islands of New Zealand.
H. barkeri,
ing placed in a large polytomy that includes this
the third Chatham Island endemic, be-
grouping. Hebe macrantha, placed with weak jack-
acks
many of the previously mentioned synapomorphies
knife support at the base of the Hebe clade.
for the group. This species has long held an isolated
or ambiguous position, being placed by Moore (in
"Grandiflo-
by Heads
(1987, 1994b). For the present we suggest its re-
tention in Hebe,
Allan, 1961) in its own grouping, Hebe
rae," and was included in Parahebe
with
1993b)
and the Leonohebe clade, monophyletic (see
which still leaves the genus,
the exclusion of Heliohebe (Garnock-Jones.
Table
TAXA OF UNCERTAIN AFFINITIES
Many of the genera and subgeneric groupings
historically recognized in the Hebe complex (e.g
Heliohebe, cushion-forming species of Chionohebe,
and some Hebe groups) are shown to be monophy-
letic in this study. Significant — are Par-
1 Allan, 1961) and
species groups within Hebe les ‚ sects. Hebe, Sub-
“(
ahebe “Group В” (of Ashwin
Annals of the
Missouri Botanical Garden
distichae, Glaucae, and the informal group “Con-
natae" of Moore (in Allan, 1961)] (see Table 1).
The
scattered, forming a grade among the New Zealand
*
species of Parahebe “Group B” are widely
hebes, and few show any close relationship to other
members within their informal group. Previous au-
thors (Ashwin in Allan, 1961; Garnock-Jones,
1976b, 1993a) have suggested the Parahebe
“Group B” species exhibit considerable morpholog-
ical similarity, but now it seems this. similarity
might derive from plesiomorphic character states
(relative to the Hebe—Heliohebe clade), from con-
vergent evolution of floral features associated with
1976b), and per-
haps from reticulate evolution in P. spathulata (dis-
self-pollination (Garnock-Jones,
cussed above). Garnock-Jones (1993a) suggested
that Parahebe “Group B” and Chionohebe should
be united, but this view receives no support from
the ITS analyses (Wagstaff & Garnock-Jones, 2000,
and herein). Wagstaff and Garnock-Jones (2000) in-
ferred that the ancestors of Parahebe and Chionoh-
—
ebe initially evolved in a montane or alpine envi-
ronment, then subsequently radiated into lowland
environments during episodes of Pleistocene gla-
ciation, Extinction probably had a more profound
effect on the basal lineages of Parahebe and Chio-
nohebe, and this process further confounds our ef-
forts to resolve relationships.
The relationships among species of Hebe remain
ter ITS
lengths are too short for us to confidently derive an
uncertain. after analyses (Figs. 3, 4). Branch
infrageneric classification from this study. It may
be that speciation and diversification in the Hebe
clade is too recent for ITS divergence to reveal its
phylogenetic pattern. Reticulate evolution might
also have clouded the molecular signal either
through diploid hybrid speciation or allopolyploidy
in this group where about 32% of the species are
polyploid.
ORIGIN,
DIVERSIFICATION, AND DISPERSAL
Two widely differing opinions have been pre-
sented regarding the age and origins of the New
Zealand hebes. Some authors (e.g., Skipworth,
1973; Heads, 1994a) have proposed a Gondwanan
origin to account for the present distribution of the
, Raven, 1973; Garnock-
1993a; Wagstaff & Garnock-Jones, 1998)
have suggested the group has arrived more recently
group, whereas others (e.g.
Jones,
in the Southern Hemisphere and that dispersal has
played an important role in shaping its distribution.
The data presented here lend support to the latter
proposition.
A primary contribution of the present work is its
assessment of relationships between Veronica and
Al-
though a close relationship between these two
Australasian members of the Hebe complex.
groups has long been assumed on morphological
grounds (e.g.. Wettstein, 1891; Cheeseman, 1925;
Raven, 1973), some authors (e.g., Hong, 1984) have
directly opposed the notion that Australasian taxa
are derived directly from within Veronica, Our anal-
ysis of rbcL clearly supports a close relationship of
Veronica to the Australasian genera, with the strict
consensus including a well-supported (100% jack-
knife) Hebe—Veronica clade (Fig. 3). Analysis of ITS
sequences shows the Australasian members to form
a clade nested within a paraphyletic Veronica. This
pattern of relationships is congruent with the notion
that the Hebe complex is an Australasian radiation
of Veronica.
Our assessment of the time frame for the origin
and radiation of the Australasian genera, and the
New Zealand hebes in particular, relies on infer-
ences from the fossil record, the geological and cli-
matic history of New Zealand, and the distributions
and ecological tolerances of extant species. For
Hebe, the earliest appearance in the fossil record is
in the Pliocene (Mildenhall, 1980), for Scrophular-
laceae it is in the mid Miocene (Tiffney, 1985), and
for the whole of the Lamiales (sensu APG, 1998) it
is in the mid Eocene (Muller, 1981). Although there
is always the possibility that older fossils will be
found, it would be inconsistent with this record to
assume that divergence, either within the Austra-
lasian Hebe complex, or between members of that
group and Veronica, occurred in Gondwanan (Cre-
taceous) or earlier times.
We acknowledge that the divergence estimates
presented in Figure 2 are crude and await further
refinement. The paucity of the fossil record and the
occurrence of undetected multiple substitutions on
Figure 7.
of trees produced by analysis of ITS sequences. Species
are regarded as alpine if
>
Details of the natural elevational range of species of New Zealand hebes, overlaid on the strict consensus
Sp
they occur predominantly in areas
above the natural tree line. The — two elevational zones are less precisely defined with, on the two main islands of
New Zealand, lowland corresponding to those areas below ca. 500 m above sea level, and montane being those between
ca. 500 m and up to ca. 1000
occur in alpine or montane environments.
1 (but below natural tree line). Members of the basal grade of New Zealand hebes all
Wagstaff et al. 53
Volume 89, Number 1
2002 Diversification of the New Zealand Hebes
sapiossa4dno әфән E
пиршәхәәцә әфән m
01р[0112 ӘДӘ та
DYINSDAJA] әфән m
p1pjoi12doup|d әфәцрлра ка
01]12 24240101417 m 3
D]DJO1]12 242940101417 ша :
siupuiajnd aqayouly) m
MUOSIUOY] 2q210u011/7) та
1491414 AGAYDAD та
Юр1141 AGAYDAD m
рцојхиәр aqayouolyy) m
pijofisuop aqayouoly) m
пиршәхәәуцә AGAYDAD та
мәрмәрира IQOYDAD E
SIJAISIADAO әфәурлра ES
pijofiui] 2QIYDADA
Montane
unordered
С] Lowland
ШШ Alpine
EA uncertain
E equivocal
Elevational Range
SUPISAUDI AGAYVDAVE mE
ADIIDAADIDI IQIYDAD f E
11104] AGAYDAD ЕЗ
pipjnyinds AqayvAdd ra
ا 2424P4D A we
1nopa 24210112] кау
pupipnp. * әфәцоцән ка
DUDIAMY әдәцо1әН{ E
DIUDAODIU m =
— 228 та
d әдә m
ТШ 2q2]] m
ay aqai m
уйно ^4 2q2H m
[1401221| 2q2]] m
pnbuido4d әфән m
Ds021u424 29H
NSUOAISIUAD 2q2]] m
юрурупиир aqa =
SIPIOVMAODM]DS 2q2H
14244Dq әфән са
pDUDAUADYIOI әфән та
DINUDYIDYO әфән a
ijo»quo[[oip әфән E
DLUISSISOLUDA 242 H на
papiuoDdo әдә] та
DSOUDAL. — әдә m
DAO
nups1440: ?^2qoH кз
Annals of the
Missouri Botanical Garden
long branches are two potentially significant sourc-
es of error, both of which could lead to inaccurate
estimates of divergence times. Though crude, the
estimates nonetheless provide intriguing compari-
sons with patterns of ecological diversification and
aspects of the geological history of New Zealand.
The geological and climatic history of New Zea-
land suggests that differentiation of the New Zea-
land hebes is likely to have occurred in the late
Tertiary. Members of the basal grade of New Zea-
land hebes, in particular the Leonohebe clade, and
the Chionohebe A and B clades (Fig. 3), are all
alpine or montane plants, most occurring in areas
above the natural tree line (Fig. 7). If the present
ecological requirements of these groups are indic-
ative of those of their past (i.e., assuming that each
lineage has not independently and recently adapted
to alpine habitats, or that each has not seen selec-
tive extinction of lowland members), it can be in-
ferred that
hebes occurred in alpine environments, with colo-
early differentiation of New Zealand
nization of the lowlands being a secondary event.
The evidence is that alpine environments have only
existed in New Zealand since the Pliocene or latest
Miocene, subsequent to the onset. of mountain
building, in what was previously relatively low-ly-
1979; Ollier, 1986).
Prerequisite in any hypothesis of a late Tertiary
ing land (Flemming,
origin for the New Zealand hebes is colonization of
New Zealand by long-distance dispersal of ances-
tral form(s). Assuming a minimum number of dis-
persals or extinctions, the topology of cladograms
derived from ITS sequences (Figs. 3, 4) suggests
that differentiation of New Zealand hebes followed
a single colonization from either Australia or Eur-
asia. This differentiation was succeeded by second-
ary dispersal from New Zealand.
As illustrated here and elsewhere (Wagstaff &
Garnock-Jones, 1998,
ITS sequences (Fig. 3) of extant species with trans-
2000) the morphology and
oceanic distributions provide evidence of the ca-
pacity of New Zealand hebes for long-distance dis-
persal. Such transoceanic distributions are seen in
four species, all of which are included in our ITS
study group, and all of PEO h are nested within the
well-supported clade of New Zealand hebes. Of
these species, Hebe — and H. salicifolia nat-
urally occur both in South America and southern
New Zealand, whereas Chionohebe ciliolata and C.
densifolia occur both in the South Island of New
Zealand and southeastern Australia. Within each of
these four species, populations separated by ocean
gaps show no apparent morphological differentia-
tion and only limited sequence divergence (Fig. 4).
This suggests that the transoceanic disjunctions in
the distributions of these species are relatively re-
cent phenomena, and given that these species and
all of their closest relatives occur within New Zea-
land, that these disjunctions are products of long-
distance dispersal of propagules from New Zealand.
Godley (1967) suggested oceanic birds as likely
vectors for the dispersal of seeds of H. elliptica (Fig.
5X) and Н. salicifolia from New Zealand to South
America. Trans-Tasman dispersal of the two alpine
Chionohebe species from New Zealand to Australia
is less intuitively explained owing to the sexual di-
morphism of one species (Delph, 1988, 1990), and
their splash cup method of seed dispersal (Gar-
nock-Jones, 1993a). The implied direction of dis-
persal from New Zealand to Australia is also
against the prevailing westerly winds but, as noted
by Wardle (1978), weather conditions sometimes
occur in which the usual direction of winds across
the
independent dispersals of Chionohebe from New
Tasman Sea is reversed. The occurrence of two
Zealand to Australia may seem unlikely, but the
alternative explanations are either an extended pe-
riod of stasis in both morphology and ITS sequenc-
es (assuming distributions produced by fragmenta-
tion of Gondwana), or widespread extinction in
Australia (assuming dispersal in the opposite di-
rection)
Apart from the dispersal prerequisite to explain
transoceanic species distributions, another six dis-
persal events from the main islands of New Zealand
are required to explain the current distribution of
Hebe (Vig. 6). Most of the postulated dispersals are
to New Zealand's
the Pleistocene-age (Sykes, 197
outlying islands, including one to
7) Kermadec lIs-
—
ands (where Hebe breviracemosa is endemic), and
three to islands of the New Zealand subantarctic
(where H. elliptica and Н. odora have populations
disjunct from those on the main islands of New
Zealand, and Н. benthamii is endemic). One dis-
persal to the Chatham Islands has been postulated
on morphological grounds (Moore in Allan, 1961;
Garnock-Jones, 1976a; Wagstaff & Garnock-Jones,
1998). This is partially supported here by analysis
of ITS
endemic species, H. chathamica and Н. dieffenba-
sequences (Fig. 4), which places two of the
chii, as sister taxa, and the third, A. barkeri, in the
polytomy that includes the branch uniting the other
two. A final dispersal, probably from the Chatham
Islands (Garnock-Jones, 1976a, 1993a), is also pos-
tulated to account for the distribution of H. rapensis
(not included in our analysis), which is endemic on
Rapa Island in French Polynesia.
The presence of Parahebe in New Guinea is dif-
ficult to explain. Here, as in the analysis of Wagstaff
and Garnock-Jones (2000), ITS
sequence data for
Volume 89, Number 1
2002
Wagstaff et al.
E aa of the New Zealand Hebes
only one New Guinean species of Parahebe were
included. That species, Р. vanderwateri, is nested
within the New Zealand hebes (Fig. 3) with 99%
jackknife support. The most parsimonious interpre-
tation of the present data (assuming a minimum
number of dispersals or extinctions) is long-dis-
tance dispersal from New Zealand to New Guinea,
as proposed by Wagstaff and Garnock-Jones (2000).
Further sequence data for New Guinean Parahebe
(of which 12 species are described) and Detzneria
(monotypic and endemic) might provide a clearer
ationships between taxa from the two
—
picture of re
areas.
Literature Cited
Albach, D. C. ‚ W. Chase. 2001. Paraphyly of Veron-
ica (Veroniceae: i. rophulariaceae): Evidence from in-
ternal transe — — d мер nces of nuclear ri-
us 11
iig й ВІ; 9-18.
Albert, Bac luni K. генно М. W. Chase. J.
R. — D. Mishler & К. C. Nixon. 1994. Fuss -
tional constraints bs rbcL — nce for land plant phy-
logen . Missouri Bot. Gard. 81: 534—567.
Allan, H н. 1961. Flora of New Ze d Vol. 1. Govern-
ment Printer. Wellin
ngto
APG. 1998 [1999]. An r classification for the fam-
ilies of flowering plants. Ann, Missouri Bot. Gard. 85:
—-553.
eem B. $ M. J. — M. n M. F. Woj-
iechowski, C. S. Campbell & J. e 1995.
The ITS region of nuclear — DNA: A valuable
source of evide — on — phylogeny. Ann. Mis-
souri Bot. Gard. & 27
Bentham, G. 1876. ^ vini Vr MN Pp. 913-980 in G.
Bentham & J. D. Hooker (editors), Genera Plantarum,
7-277.
Vol. 2. Reeve, London.
Bremer, К. & M. Н. С. Gustafsson. 1997. East Gondwana
ancestry of the sunflower alliance a families. Proc
Natl. Ac p Sci. U.S.A. 94: 9188-919
Briggs, В. G. & К. Ehrendorfer. 1992. 4 revision of the
Austr hs species of а and Parahebe (Scro-
phulariaceae). Telopea 5: 241—2t
Carlquist, S. 1974. Island Biology. ИР" Univ. Press,
Y
Ne rk.
Chase, M. W., D Soltis. R. G. Olmstead, D. Morgan,
D. H. Les, B. D Mishler. M. R. Duvall, R Price,
H. G. Hills, Y.-L. Qiu. K. A. Kron, J. H. Rettig, E
Conti, J. e nns J. n. Manhart, K. J. Sytsma, H. J.
Michaels. W. J. Kress, К. G. Karol, W. D. Clark, М.
Hedrén, A S. Gaut, R. K. Jansen, K.-J. Kim, C. F.
s, Q.-
Wimpee, J. F. Smith, G. R. Furnier, S. H. Straus
Xiang, G. M. Plunkett, P. S. Soltis, S
S. К. Williams, P. A. oo С. t Quinn, L.
E. Golenberg, G. H. Learn, ge Graham, 5. С.
Barrett, S. E dee no UR 1993. Phylo-
genelies of seed. plants: An ees of nucleotide se-
quences — the plastid к gene rbcL. Ann. Missouri Bot.
520-50
. M. Swensen,
E. Eg не
Н.
Gard. 80: 528-
Cheeseman, T F. pel — of the New Zealand Flora,
p 5 rnment Printer, Wellington.
Deg 1988. The Evolution and Maintenance of
Gender еей in New Zealand Hebe (Scrophular-
iaceae). Unpublished Ph.D. Thesis, University of Can-
— ыр һигеһ
The А of gender dimorphism in
New а Hebe (Scrophulariaceae). Evol. Trends PI.
4: 85-97,
Doyle, J. J. € J. L. Doyle. 1987. A rapid DNA isolation
procedure for sal pousse of fresh m tissue. Phy-
tochem. Bull. Soc. Ames 19: 11-15
"ama, J. S., V. E Albert t. M. Källersjö, D. Lipscomb &
. Kluge. 1996. Parsimony nk outperforms
s Il:
— — of New
Auckland.
Brown)
а!
neighbor-joining 2 — stic
Flemming, C. A. The
кы and Its Ж. Pu Капа Univ. Press.
Garnock-Jones, P. 976a. Hebe rapensis (F.
Garn-Jones — nov. and its relationships. New Zea-
land J. Bot. e 79-83.
. 1976b. Breeding systems and po " New
Zealand Paral (Se rophulariac eae). New Zealand J.
Bot. 14: 291—
em За. D idee ny of the Hebe e —
ET roniceae). Austra „ Syet, Bot
J3b. Heliohebe (S (Ser
a new genus í segregated from Hebe. New Ze b J. Bot.
(Se —
: 457-479
roniceae \.
lariaceae:
31: 323-3:
Godley, К. ñ 1967. Widely i ta г species, land
bridges Үк continental drift. ге 214: 74—75.
Heads, М. J. P New names in — Zealand Scrophu-
Bot. . Otago News 5: 4—11.
ende 'eae.
1994a. E Ау are iter и and taxonomy
in the Hebe sum on pee à eae). ull.
Natl. Hist. Nat., B, Adansonia 16:
994b. biogeographic
ae). Bol. J
FE |
\ t
Hershkovitz, M. A. & E. A. 4
теме w of Parahebe
. 115: 65-89.
. Conservation
patterns in angiosperm rDNA sequences. Nucl. Acids
Res. 24: qe d
чер P. K.. М. H. Holmgren & L. C. Barnett. 1990.
Index РЕ ЯВ: The dh of the World,
8th ed. iq
Hong, D. 1984. Taxonomy and evolution of the Veroniceae
(Sc — eae) with s special reference to palynology.
Opera Bot. 75: 5—00.
oe M., J. S. Farris, M. W. Chase,
ay, C. J. Humphries, G. Petersen,
ы 1998. Анесов parsimony jackknife i
ysis of 2538 rbcL DNA es reveals support for
major clades of green — land — seed plants
and — plants. Pl. Syst. Evol. 213: 259-287.
Lin, C. M., Z. Q. Liu & 5. D "King. 1986. P deii
roplast genome: Correlation between the DNA sequenc-
es RÀ the isoelectric focusing pattern of the LS of rub-
lec. Biol. 6: 81—87.
Liu Tih- im & C. L. Schardl. 1994.
quence in inte mal transcribed spacer l of plant nue * ar
rRNA genes. PI. . Biol. 26: 775-778.
Mildenhall, D. С. ‘oa. Ne aw Zealand Late Cretaceous and
Cenozoic plant biogeography: A $e Palaeo-
Part 1,
О.
Veg.
B. Bremer, M. F.
K
Seberg &
sequence
-
A conserved s
:eogr., Palaeoclimatol., Palaeoecol. 31: 197-23:
Muller, J. 19 Fossil pollen rec wu at extant angio-
sperms. Bot. Kev. (London) 47: 1—
Ollier, C. D. 1986. F origin of id xoc in Aus-
tralasia. Pp. 3-25 in B. A. Barlow (editor). Flora and
Fauna of Alpine ba. Ages and um CSIRO
1 Melbourne.
Olmstead, Н. G. & P. A. Reeves. 1995. Polyphyletic origin
of the Sc lara eae: Evidence from je y ndhk
Ann. Missouri Bot. Gard. 82: —193.
Australia,
sequences.
Annals of the
Missouri Botanical Garden
„Н. J. Michaels, К. M. Scott & J. D. Palmer. 1992,
Monophyly of the Asteridae and identification of their
major — inferred from DNA sequences of rbcL.
Ann. Missouri Bot. Gard. 79: 249-265.
—- .. M. Scott & J. D. Palmer. 1993, А
parsimony analysis of the Asteridae sensu lato based on
rbcL sequences. Ann. Missouri a — 80: 700—722.
— ———-, C. W. DePamphilis, A. D. . N. D. Young &
Р. A. Reeves. 2001. Disinte ie of dhe Scrophulari-
aceae. — i — 88: 163-212.
Pole, M. 19 > New Zealand flora—Entirely long-
distance ай Г . Biogeogr. 21: 625-635.
Haven, P. H. 1973. Ev olution of the subalpine and e
ри groups in New Zealand. New Zealand J. I {
—2(
Re a J. L., W. S. Judd € R. G. Olmstead. 1999, Pro-
posal to conserve the name Antirrhinaceae and with :
superconservation proposal to — Анпесе
over Plantaginaceae (Magnoliophyta). Taxon 48: 182.
Continental drift and Ne ^w A 'ü-
. New с and J. Geogr. 57: 1-1:
1998. PAUP*: Phyloge — — Us-
ing Seed ( кыз d methods). Vers.
Sunderland, Ma assachus
Sykes, W. 7. Kermadec Island Flora, an Annotated
Checklist. Ne. W Ze aland Department of Scientific and
4. Sinauer,
Industrial Research, Bull.
Wellington.
Eve J. D. . Gibson, F. Plewniak, F. Jeanmougin
& D. нон 1997. The CLUSTAL_X windows in-
s F lexible strategies for multiple sequence align-
ment aided by quality analysis tools. Nucl. Acids Res.
25: 4870-4882.
Tiffney, B. H. 1985. Perspectives on the origin of the flo-
ristic similarity between Eastern Asia and
Americ а, J. Arnold Arbor. 66: 73-94.
Wagstaff, S. J. & P. J. — Jones. 1998. Evolution and
bioge pla of the Hebe complex (Se WAV e x i
inferred from Is. sequences. New Zealand J. Bot.
ime
219. Government Printer,
eastern North
. 2000. Patterns of diversification in
Chionohebe and ‘Porahabe (Se ра inferred
from 1 F s — ices. New Zealand J. Bo 389-407.
— Wardle. 1999, Whipcord he be — mat⸗
ics, dian ecology and evolution. New Zealand J.
Bot. : 17-3
Wardle, ÉS 1978. — of the New Zealand mountain flo-
ra, with special reference to trans-Tasman relationships.
New Zealand J. Bot. 16: 535—550.
wes in, R. von. 1891. Sc — eae. Pp. 39-107 in
. Engler & K. Prantl (editors), Die natürlichen Pflan-
ze ue т, Vol. 4, part За. W. Engelmann, Leipzig.
Wagstaff et al.
Volume 89, Number 1
2002
tion of the New Zealand Hebes
IVersifica
D
OLOPEOAV
€c89c0AV
LOOPEOAV
Fe89c0AV
1891 UI
029€c LIV
889111
c29tc LIV
€£29t£cLTIV
OVOOZZA V
T8ELEOAV
I9€2£04V
6t06cc V
8t06cc1V
CLELEOAV
OVSYEOAY
LEOGECAYV
LEOOZZA V
(0007) souof
-yoouresy) Y JLISSEN
(8661) *»uof
-YOOUIRL) Y JJEISBE үү
(8661) sauof
-YOOUIE у jjersaew
(0007) Te 19 peojsui[()
(000) seuof
-yoourey) Y Jase
(000) səuof
-ҳоошех) Y jeissew
(8661) sauof
-ҳоошвх) Y JJPISSPNM
1oded Бї]
(000Z) səuof
-JJoue Y Jessen,
(000) sauof
-yoouresy) Y JJPISBE ү
(0007) “Je 19 peaswjo
—
2661) [E 19 peaisui[()
(0002) “Te 19 peairsuio
—
(£661) ‘Je 19 peoisuio
(0007) [€ 19 peaisumo
(0002) “18 19 Рәш)
corels HHD
98tcl€ HH?
69tcl€ YH
C8PelS ЧНО
9c696t HH)
PpovelS HHO
CBL9Eo HHO
8esoor YH)
189806 YH)
69619
ео
985819
€6/TSS
9ECOLD
OF tO
окар uapae) [p1uourLrodx*g цәлвәв
-ƏH atRopuey рәли) Сү [VHLISOV
SOI T6
fpisany пәрлесу үешәшиәахУ цәлвәѕ
-ƏY әвәриет рәли) "VFTVHLISOV
08122
uosqo(q iL "y FIUEWSE] "VIVA LSOV
6821'8 sau0f-190U.1D1) f a
‘upusaH а ‘Yed 904 '(INV'IVAZ WAN
£6/[€€ SIYÁS Y M *jurod 3117 juuns
чпцну “YA “UOSTON (IN VIVHZ MAN
0666 “Чоон
d Jseurdy tmuoyoH, CUNVIVAZ WAN
FUOLSUAY сри
OOLE ye кәү әшчүү OISNLISON әзиру
шер “SSP FIOOUWON CVETIVHISTV
£6 “dy zz
“ploy у шох] `Ч 2P4 d "9sueM
319qp "puepsow "(CN VTVHZ MAN
0029 uosduig Y TN 8304) uo nue H
‘puowog пә ^erueuisep үүнү
(ASA) Апола veugrpu] “HOY xo (CCR
иәрдорү M 7) SISS sipiduing op Mo
(NNOD) ZSIZ YM
(OVA) ZEr06 sipiydumqgop M 7)
(ASA) URSIYOI N Jo AJISIOA
чий “HOY хә P0706 YA WD LP M 7
(DVd) £o ro sipidumgop M 7)
(эма) ££706 sipidungop M 7)
yey (14 Y) pinyofied niuomao(q
р а
y s33LIg 9 “Y (Три) veu ю1уиәлмә(]
puaiq Ņ s33ug
су ч (swoipuy) рирциәлмәр DYUINLA]
Три p1oyfipupas 018и11]07)
puaiq4 y sssug
с) 79 (ueueqong) znuosuiojy] 242401017)
pusiqq Y s3311g
2) ч UJ "100H) siubura nd aqayouoly’)
"pualyg Ņ Sodlig
299 Сөп У) pijofisuop aqayouoly’)
“‘puslyy y БИСЕ
9 Ч (TENA У) pijofisuap aqayouory’)
‘puslyy ow 833119
“үчү? чоон) 01010119 aqeyouoly’)
puaiq y s33LIg
°“) 9 CJ 00H) 217701712 aqauouonj?)
"ү pnbijgo 2u0]217)
qeq
‘puss using pypAydosajay 2421111107)
"Чоң "p ^g pupiuoadD2 ndoong
“т snou шпицітиү
“‘yWueg suaosaqnd piuojasuy
зло], snjpsnd snyjupiyduy
“1904
Sil
UOISSIII® que gus)
UOH BIO
ainje19jr]
UOISSIDIB
UnLreq. 9H
UOISSIDIR
U3p.re4)
a21n0s/411[P20] [еши
sato9da
"sip шол
I[QPAVLI] ol UOT]EULIOJUT UOot]29[[09 3]3]duioo чим (ишпүоә ципо 395) 19qUunNUu UOISSIDIP anbrun В ӘАТӘдӘ1 WALI GAY HH) ә ul suauttoedg “(0661 “Те 19 US UOH) wundoq H хәри]
моо) 'sasaujuaed ut "SUOT]EIASIQ(QE Untcrec.r9 H ‘Арп INO Ul p2pn[our әвәәвиеүпцЧоләс jo soroads ou 10] sSIoquinu UODISSIDIR yurquos) pue UOTPULIOJUT 1IYINOA
‘| xtpueddy
Annals of the
Missouri Botanical Garden
LLOPEOAV
VIOTCOAV
800V E0AV
ELOPEOAV
LE89COAV
ScctcIAV
606 LOT
COELEOAV
CCOTCOAV
BLELEOAV
P8ELEOAV
60€ LE0AV
[S8LEDAV
LLELEOAV
L8ELEOAV
106774V
VLELEOAV
£91690.1V
1916904 V
ELELEDAV
(8661) seuof
-Pouw у Jese
1aded siy}
(8661) souo[
-ADOUIBO Y JJe1se M
(8661) souof
- oou.) » jessew
(8661) səuof
-ҳэошв‹) A еар
taded siy
(8661) souof
-AYOUIBL) Y jersey
(8661 ) səuof
-ҳәоше‹) Y JPEISBPAY
(0007) S9UOf-yooure+)
y резем ded siy}
(8661 ) səuof
-JPOURA Y рер
(6661)
3[pI£A Y JJesae ү
(6661)
PILA Y Jesse yy
(8661) səuof
-YOOUIBL) Y рер
(0007) 7° 19 peaisumo
(0007) [e 19 peaijsuio)
(€661) "|? 19 pearsum()
26891 ALIJA
COPIE HHO
OotbtcI€ HHO
8€tcI€ YHD
trbcIC HHO
pLvVclS HHO
cLicI€ YHD
bevels HHO
OSO89P HHO
9бћсІс ННЭ
ESPECIES HHO
LOVISES YHD
ВРІ HHO
££6€1:)
2£9819
460/19
Калы
L6/cóc
16/60
SLOFID
190919
900919
v6/99€
PELIS
APAPHPW `N “PI
SSEJIE') ‘ривер puPpi[e 4 ‘VNILNAOHV
SIOFI uou) ^M q eoiexureN
‘ривер шефецо “INVIVAZ WAN
6861 ?9(qp SI ‘uong `y ‘3109
Pee y Anduaau "(IN VTIVAZ MAN
P86] uossav]) “Y
g seuof-3o0u40*) ^f. q “ред suoi]
“OW PUE Pny quinos “INVIVAZ MAN
£881 souof-yoouane) “fy KA
әрпинәсу "puepqinos “INVIVAZ MAN
66'p'€] Louooy “Gq ۸2|
[PA APOJIABH ‘UuoS[3N INV IVAZ MAN
6661
кәиоГ-уәоили) `f J “UDUIIH 4 “URIS
921027) “YSNOLOGHEW “INVIVAZ MAN
swni
'V d Ot£gc[ чәл) y q eoresureN
‘Spue[S] шефеҷо (NVITVMZ WAN
066! AON LE ?2uDM “Y “Y “UH Yew
-994 ^pue[s] [[oqdure?) INV IVAZ WAN
€g-[c[ uou)
Y dg pue[sp uei) “UNWIVAZ AIN
CBEOZ хорор og muas
“oy лих AMG RO INV IVAZ MAN
v661 “eq, “мохаш
A PIDA q “ndyeyey oye] Ieu
1941. 3AM “PUBTYIMOS "(IN VIVAZ WAN
gg TE A9U00Y (| Je] лзәрүпоң
"uos[oN ISIMULION. (IN V'TVHZ WAN
F£'06 sipjdumgop MD
Jayonoa ou тоту "I
uwouxun
[[puuaq (718104 *£)) nondijJo әфән
uelly
y әчАрхәогу (yug) ry2pquaffoip əqəH
wy Y
JUABADO?) (J *NooH) sapiossaudno əqəH
ƏS) NUDÍLLIO) 292
ue[[V w eu&e
1907) (чешәзәәцг)) DUDEUADYI0I aqap
uelly
Y әчАвңәогу (J ‘YooH) DIDIONII әфән
uelly Y
JUÁBADO?) (ueueyong) 1uDuas2242 әдә
"IV
JUÁLADO?) (ueueqong) DINUDYIDYI 29H
uP[[V
y QUARYIO') (у *400H) 11upquoq әрән
JUABADO?) (QUARYIOD) 14240Q ƏQƏH
uP||V Y әцАрР
4907) (ASU cg 7f) 272uo0418uun 2q3H
ue]
-|y y эч $207) (910194) njnjnuun aqa
JUABADO?) (21124) SUDIIG]D әфән
“XY, »sopid jonas)
"] DIJOfIp10) Diunqnqoj^)
“| paindind sippnziqq
71291
SLI
UOISS32)P yuequo«)
uot 1815
21nje13]r]
UOISSad0R
unuıeq12H
UOISSIIIE
пәрлес)
әәлпозулүүрәо] euB
saraadç
‘penunuoyy “р xipuaddy
59
Wagstaff et al.
Volume 89, Number 1
2002
tion of the New Zealand Hebes
iversitica
D
— . SS8Tt0AV
cV 06ccAV
9910690.1 V
— . vy€8TV€£0AV
LIOVEOAY — 88€L€01V
OIOTEOAV —-.][ L6£2 £04 V
— . €£€9FCOAV
9€T0690,1 V
00€ 2 €04V
0916904 V
C[OTCOAV —. €E8ELEOAV
698€2£04V
— . EO€LEDAV
1oded si
(0007) souof
-POUA у JSB
(6661)
[DIEM 5 PRISA
1oded siy}
(8661) seuof
-4ooude4) Y увів
(8661) s9uof
-YyooulR4) Y етае
1oded sip
(6661)
эрюм Y jisse ^
(8661) s2uof
-YOOUIPL) Y JLS.
(6661)
PEN Y уез ү
(9661) sauof
-yoourrey Y JJPISBE ү
(8661) Souof
-YDOUIP<) Y це yy
(8661) souof
- YOUR) y PRISA.
86719 YH)
I9tcI€ HHO
LVVGIS HHO
IprelS ЧНО
Реге ЧНО
8971 YHD
O8belS ЧНО
18966 HH)
O0€tcI€ HHO
Oorels ЧНО
62YcI€ ЧНЭ
T8
=>
TIE ЧНО
906219
4802819
881219
08219
218219
0L£819
188819
16/981
68 661 Чош)
Г `хә Suape puswuodxy qirase]
3IPopue'[ PARADA 'ANVIVAZ MAN
c06 1 SUOS -Y30U104) T d
Yoo) OLAS, "pue[uinos "(IN VIVAZ WAN
Q9'CC AJUOOY “(| "S19
-шо 1N CAInqaoque?) IN VIVAZ WAN
6861
ириәә} q AIQS1NN [994 “HA шол) «¡eu
-IBLIO *uopae«) puswnodxy цолгәвә
aleopuey P3۸11") “Q(NWIVAZ WAN
0661 ?9(] t sauof -20u4D*)
T d “Stesiny [994 WW ONVTIVHZ WAN
681121 uDuaapg d UP
-189 OURO Y9INYOISLIYA) шолу A[[eu
-13110 *uapaes үрішәшиәахә yo.eosoy
элворив” рәли) (TINV'[VMZ WAN
60'€6 fisa ^f су uuo “Y
nejd WORT “UOSTON CUNVIVAZ WAN
£0'€o Ліха f s
“OEL VW | апав») XINVIIVUWZ WAN
xel ady nq
d V (eBuey әшцепу jo ands) osury
OYOWRDY “ABE SMEH “UNWIVAZ WAN
cg61 $9UO[-Y90U1D2) "f. 4| SUY
nuire pueipnnos (INVTVMAZ MAN
OLEVOLB
вшәдивә) *suopaee) oruejog “VIV MLS IV
10°66 ffoiszog ^f ^s
`оҹеләр 1N AMARO "(IN V'TVHZ MAN
68'6c[ YP
1994 “HA Чо «¡eu
-15110 "uapaez [Piuour1adxa цәолвәвә}
элроривт POP ALN “(NV IWVAZ WAN
-uo9H d *хләвп!\
ug[[V
y JUÁBADO”) (учоон) sapioajound әдә}
черү
у eudRyoory (ивирцәп ) mud 997
9100] ч “I
(uejjy Y ou&eo07)) vsowvnonod oqop]
ue]
-[V Y vudeyoor (үү) p40jfri40d әдәр
QUARYIOT) (у 00H) D40po aqoH]
ueT|V
y ousroo?) (J 00H) DYJUDLIDUL әдә
ue[[Y
Y OUARYOOT) ("у YOOH) DYJUDIIDUL 99H]
sopioipodoo4] ‘dsqns uelly Y
JUABADOS) (J Чооң) sopiorpodoot] əqəH
PEN y reise
(OSUa]O")) хушхат "dsqns 11401994 99H
1101994 ‘dsqns ue]
-[V Y 2useNo07) (р 00H) 11103994 әдә
м
DSOUIOS әфән
перу
y QUAPY¬207) ("у Чооң) рәрмәраә 991
ouARyoory (14 "M
Ццәччә,] (18104 *£)) Didi] 99H
qoe SLI
UOISSIIIR quegquo: )
поро
әлпрлә1”]
UOTSSIIIR
WNLIEGLO LY
UOISSIIIB
Uo pe)
QIINOS/AT [RIO] JEU)
sotoada
‘panuyuoyy `] xipuadd y
Annals of the
60
Missouri Botanical Garden
(0007) ѕэчо[
OI6I sauof
-YIOULDY "f q 72247) 214] 19dd 7] *urej
SpeəH
АЧ -Pour Y geiszew c€tcI€ HHD Z68819 -uno[y 2144 pue[qinog “INVIVAZ MAN (S9UO[-"UIBL)) $17Á 151421 9QIYDID
шоу "Ww M 6802
sauof-320u404) "f а ‘ш OPEL “Yea
(0007) sauof IƏMOJ, JO Y Ənbar ш ue] әлоде *ogue “ALLO
PVOOGCAV -Y 20U14) у Jese TOIOZT HHO nuper "puepqnos "(NVTVAZ WAN CHOW CA (19 74 N) 26949 292190404
(S661)
єє 894929M Y peoajsui() 261-66 $2122 YP projsui) "I suungqma sumddig
(8661) seuof | 9g'c'cc ÁNO] "7 *ә}
600PEDAY — О8Є/ЄО4ЯУ -YOOUIRL) Y Jesse 6StelS HHO C8291 -J4 UN "usnodoqpre w "INVIVAZ MAN Sauof-‘ures) (J ооң) mona 29/01/27
(0007) seuof 286'8 tono ^g “peoy unns *e[nsur sauof
€r06ccAV YOURS) IY greissew 28VcI€ HHO 90819 — -uaq sxueg “Amquajuer) UNV'IVAZ MAN -URS (moe) nupipnpap] aqayorjay
(8661) sauof 996T TLE 4249 q “a souof
6LELEOAV -YOOUIRL) x JBE O09tcI€ HHO €LIZI) -MY ә ‘Ҷапоо рер “UNVIVAZ MAN UIE) (PNW 74) PuUDIIINY 29240112
8981 sauof-20u40*) `f q “yung ue[[V
8S8PEOAV 1oded sip €€9CO9p HH) 4265/19 ndejesune|y "uospN “INVIVAZ MAN y AUÁRADO?) (J *400H) рѕоәтиләл ән
"88 6% ?!pinq "A “Y “15207 uelly Y
LC8TPEOAV 1oded siu O8ÞPZSI HHO 111/19 `по5]әМ ISOMYHON "(IN V. TVAZ MAN = 9u&exoo7) (ueuasa24)) HUOSUNO] әдән
LOBPEOAV 680102 aon] а "y `чешәѕәәц) | uwy
998V E0 AV jaded sip TEVZIS HHO 911819 YA подив) “INVIVAZ MAN Y SUABADO)) (Y *400H) 24211801121 әдән
[ZII 112]
COBLEDAV -Dg `[ “Usa "M туо Muesuneja UBITV
POSVEDAV toded sry} LEPZIS HHO 180719 pueppony YON “INVIVAZ MAN Y eudeyoory (uung *y) Dso12ads 9дәН
(6661) 990196 лон Cg `[ “owuey | uelly Y
6LOVEOAV = СОТ6ООЯУ эрге Y JISIM S'vtcI€ ЧНЭ 20999 '"sseq sxpe[ “AINGIONUPD (QN V'TVAZ MAN 9 9useoo») (y ^xooH) 59р101и1021]05 әдә}
(8661) səuof OT'S6 fisan f S
S8ELEOAV -YOOUIR) Y Jesse 99PEIS HH) "Ned ерү Cuos[oN. "(IN V'TVAZ MAN [[puuaq (715104 4) n1Jofiorpos aqa
(8661) səuof
SLOFEOAV 98BELEOAV -Ҹэоше‹) зу geiszew 86891 ПІЛУ М 6661 uoixng ^M PEPY ues “A TIHD [[puuaq ("15104 *£)) DIJOfIDDS əqəH
0616 чош”) ‘f *nqnuoen uosulou [
9C8T€0AV taded sip 8c9cI€ ЧНЭ 699819 -deL W `Ҷапогодеу “UNVIVAZ MAN 'S [y човЧш “9 DUISSISOUIDA 99H
(6661) COE OE ?]p4v4 d “9APT F10۸ | uelly Y
8€T6904V PIE y Jessen 8LPETS ЧНО H8€6/L0I -EW unos “PURTYINOS “(INVIVAZ MAN — 9useoo7) (ueuasaay)) DNbuidosd 99H
"|2q4 SLI поцеџә UOISS322P UOISS322P 321nos/A]t[eoo[ теш) sardadg
UOISS9200 эшо) ənə r] шпиефәң uapaez)
'penuguo? “| xipuaddy
Wagstaff et al.
Volume 89, Number 1
2002
Diversification of the New Zealand Hebes
OcOT £O AV
I€06cc AV
198, €0 AV
0€06cc.4V
€6t7€04V
8r06cc4AV
LVO6CCAV
O98TEOAV
LTOOCCV
9F06ECAV
6S8VEDAY
POELEOAV
9662 €04 V
(0007) зәчо[
-yooures) у JJPISSEN
ләава siy}
(0007) sauof
-ҳәоошве) Y руеіѕаем
(8661) seuof
-yoouie4) у JASSEN
(0007) Sauof
-A90UIBA) Y jpessew
(0007) sauof
-YOOUIR) Y JLISSEN
1oded si
(0007) seuof
-yvouresy » jesse
(0007) seuof
-ҳоошвх) y Jase
Joded sup
(8661) sauof
-YOOUIR) Y JISE
(8661) sauof
-YOOUIR) Y JJEISSB AY
09891 ALIAM
9c9cI€ YHD
OPPIE ЧНО
EOS HHO
LpCOLT HH)
LOTECIC ЧНО
clc60c HHO
9CcpcIS HHO
OtreIc HHO
SO/LYI
#66219
16/28 1
688119
06/T9c
077 Sauof-yoou
-1D4) T d WU 009] “peor Jo do] 1e saul
-p[Inq oprsaq “ple Bis вол], *nqodeny
"aw ‘nejd NULA CIN VTVHIZ MAN
POPE sauof -320u4D-) "D d *AYISIDATU(]
eop рәли (INV'IVHZ MAN
c6 Sl
uosdung ‘N ‘ndneyey eye] API MA
‘uoyssury “PUB]YIMOS "(IN VTVHAZ MAN
609€ [T ?manqd M Сү 1819294
"uA Anque YHON (NVTVAIZ WAN
COST uosuwuof “[ 422
- cg SAA 9 9 neg PEN sure
еи ənjg ‘speq YHOMJUI AY Jo S Wy
) jnoxoor| suoye pop MOTE xoed
uy o uy | SILA unos MON "VITTVHLIS OV
1661 `Чә4 cc ??"(1 d V лоор
T dog ang») SAQUIRT) “A9][LA
xpo[oaep] nqog) "(IN VIVAZ AIN
QI pg souo[-joouan*) ‘f 44 nu
-odeny jy "uoisur]ey CINVIVAZ MAN
90°66 кар fF су Kuua Y
тро HPe[ CUOS[T2N. (IN VTVHZ. MAN
TOO] “Чү gZ vouno] 7] '22uDM '(] ^g
"p|[ouy«ous MOQUIBY Jo Ay urseq ‘SULY
pneury S ‘UOS[9N “G(NWIVAZ MAN
EOPE
souof-Yyooulny `f а "pM sseq 10u
“TEM unodeuepy “03810 (INVTVAZ WAN
SOLOL PSL “PUM cy d FAY
BURA, “Ҷапоо ер "(IN V'TV4IZ MAN
0661 ny
“s9UOf[ -90UIDO ‘f d *uopu&] axe
Jo pu? S ‘mque "(INVTVAZ WAN
MIO
“Y YH cM Счїшәя) vi nyiods oqoupang
ALO
"a MOM Cug) орту oqojupang]
»
'd
AIO “A "Y № (иоѕшоці, ‘S "f
uosdurig :0)) njnjoijadounjd aq. png
"MIO
HOHEM (F ооң) 117794] 290040]
‘pus
1 y s3ug +9 ^g opydoyn) oqoupang
AMO
“Y HUM (J YooH) ofi] eqoyoing
A10
‘YM CAJA) DUDILIYOOY 242410
UIMYSY юлодәр 2424010
АЦО
YOM Счшәң) пиршәѕәәцә 9QIYDAD Y
IDIIDLIDIDI “dsqns АЦО "d
Y MÁ 3510 4 5) 9D]2DA44D]D2 9Q9YDID]
sauo0f-* UIE“)
пили ‘dsqns IDIIDLIDIDO 9Q9YDAD]
АЦО 'd "d Eu SUIISIUDI 9094010]
PY SLI
UOISS322E YUE ua)
uone
31n]e19]rT
UOISSIIOR
ш neq H
UOISSIIIE
пәр1їєс)
әздповулуүвдо] TRULSLI()
satoadg
'penunuo? сү xipuadd y
Annals of the
62
Missouri Botanical Garden
€66cItAV
TcO€I€AV
PCOPEOAV
66661 EAV
900ETEAV
900€ LEAV
FO0tICAV
£00€ LE AV
COBPEOAV
OTIO£ICAV
000€ LEAVY
CZOVEDAV
COBVEDAV
IcOTEOAV — LOELEOAV
CCOELEAV
— 157981
ECOPEOAV
c€06cc.1V
OLELEOAV
(1007) эеэ Y Yeqry
(1002) 9Sty) Y ҷову
1aded su
(100g) esky Y uoeqry
(1007) эеэ) Y uoeqiy
(1007) 9s249 Y uoeqry
(1007) 29884) Y uoeqiy
(1002) sey 9 Чорду
1oded siy}
(1007) эѕецэ Y ҷову
(1007) 25U) Y qoeqiy
taded sy}
(8661) səuof
-J POUL) Y JJPISEN
(1007) ase Y quoeqry
(£661)
зәләә} Y peasujo
(0007) səuof
-YOOUIR) y резе
(8661) səuof
-YOOUIR’) у jesze
[ЄВС HHO
“моя “u's
22971S YHO
NNO us
€c9cI€ HHO
S9bZIS MHD
NNOg “u's
c000€€ HH)
crvelS HHO
94AN
cVEl-BLol MAM
L1€0€-696[ MAY
PSSPl-EZ6L MAM
uuog uəpes) "og
98€T1-026T. MAM
Р6/281
192-9661 MAM
6/981
(IM) SOFI NNAS 7] “OMAN “XAMYNL
(М) S8€2 ОМИ
имо Zune 4 'NIVLIN LV AMO
F861 AMI £z sqoo4g q
є "]ZIuas07) orq uqe) “Boy “LTV LI
(М) ZIZ OMN “мэм ОЯН “UMouyuy
(М) Z768 ЭМИ ™aY OAH “AOAAHD
86/Є12 194951] “VANAM
uuog чэрв) "log "uMou»xur
(М) OIL
JAN AIAHPH "SO :pue[sioH "AVAHON
8107098
BLI9{UPT) *suapae) oruejog шолу
A[[PUISLIO “USpIeo) [ерушәшиәаху qoıeas
-ƏY әлеәрицет PARADA ‘VIIVHLSNY
(М) OIL DAN 297242) q f MALE A
1900H 23219q0x20N, suey "VIHISTV
uuog UJpJe«) "log "uMouyxur
OLEPOLB
"wLHOque-) *suopae«) oruejog шолу
AT[PUIBLIO “Uapiey [еұшәшиәаху qoaeas
-9H әлгәрцет PARANA "VI'TVHISTV
6076 MISA ^f "s 4991)
suey cnqiojue?) “(IN VIVAZ MOAN
uuog чэрве) "lod "uMouxur
"HS p»ojsumn()
8661 “des ZI
"uop]xng ^M L] рәлојоә-івәрә лвәи ‘u
OSIP “EW ‘IEA u212N 'VANINS MAN
[9[c Seuof-yoouswy "f 4 “uu Оф]
`ишшп» 1e2u ado[s y "uossuua[ IW
SIW эглте‹) “PURTYINOG “(INVIVAZ MAN
"uOJOM SISUd}]0 DITUOLIA
“I sijpuraffo DITUOLA
"p spumo varuosa
[PA DAYIDISOLIDUL DIVUOLIA
"UIS Y "и 291078 юәтиолә\
‘YOO psompup]2 3140424
"| Psopmonnaf 02100424
S. (рә DUD Yo 0.1 0194
ug "d DUIJA]DI 0.110494
"[ SOP101P1)J9) n21u042
T Saproipijjeq Da 1
"ү DIDUISND n21400494
‘Ig y 077540 DJ1U0424
"1 оәотрр-ғ1одтир DJ1u0424
zid() што uonjopunstjopnosqq
"| P1DJoo2un] оошту
пә
-AO7Y d (Пэччәд) MaJDNIPUDA 9(Q91[D4Dd
MIO 7H H MA PPH oqoupapg
PH SLI
UOTSSIIIR YUE f] ud‘)
uone
31nje1o1rT
UOISS9)2)8
uin LIEq.9H
UOISS32)P
U3p.e*)
321nos/Aj[eoo] [еши
sateds
'penunuo) `p xipuaddy
63
Wagstaff et al.
Volume 89, Number 1
2002
Diversification of the New Zealand Hebes
€cOt L€AV
GIOPEOAV — 80ELOEAV
LLO€ Leavy
618980 AV
COOE Lt. V
CSOT
ECOPEOAVY 8L89€0OAV
(1002) e8eyo 29 HOPqIV
(8661) Sauof
-ҳоошвх) Y jpeiszew
(100c) 9sty) Y Чәчү
1oded Бї]
(1007) 2884) Y uoeqiv
(S661)
89429291] Y ркәвщ()
iaded sup
NNO Us
сурст HHO
NNO “U's
£€£881€ HHO
cerels ЧНО
uuot] U9p.re*) "lod
|9joxrounie?) *uojuewN “VIH IS AV
60'P6 S2UO[ -Yy90U1D%) Г J око
f `$ ‘uepıes) [RJUGULIEd XY Youresey
a129pue] ppAnIn) “ONVIVAZ MAN
uuog uapır<) лоя "uwouxur
cce ffs f s ious
IAB] BUTE әлциәлрр ‘ALSO әлпүрү
SOUAOI) “AINGIOIURT) (IN V'IVAIZ. WAN
(NMA) 80pT nnas "Т SAN unouxur
PP Ito proisu)
TO'T6 [pisso T CS сәличәлре U9")
YURASAY 39U31 pue [PAN] noL y
"upoour] MG RO INV IVAZ WAN
"boef popup DIUI/MAL
auuag (VJ) tunonaqis um4]spoiuo4o
II d CI 1110 140124
"boef 0170/121111 02140421
"| viJof ¡Adios no1u0424
“SIA sopioloanips D21U0424
“1104 DJIs49d. DIVUOLOA
но, D3is4od DJ1U0494
1294 SLI
U0ISSIIIE ЧЕ] пә“)
uon
31n]e191r]
UOTSSIDIIR
шивә p
UOISSIIIR
Uope«)
321nos/A]i[poo| [еши
әтәч
‘ponuyuoy | xipuodd y
PHYLOGENETIC
RECONSTRUCTION OF THE
NEOTROPICAL FAMILY
QUIINACEAE
(MALPIGHIALES) BASED ON
MORPHOLOGY WITH
REMARKS ON THE
EVOLUTION OF AN
ANDRODIOECIOUS SEX
DISTRIBUTION!
Julio V. Schneider,
Georg Zizka?
Ulf Swenson,? and
ABSTRACT
Based on morphology, a cladistic analysis of the Neotropical family Quiinaceae (Malpighiales) was performed to
generate a hypothesis of the ر netic relationships within the family.
'hnaceae were used as
all genera, apart from Lacunaria,
species of Oc
are monophyletic
The monotypic Medusagynaceae and four
teroup. Sing equal weights, ins Quiinac eae find strong jackknife support and
. Lacunaria receive
support only after successive weighting. The
aberrant species Lacunaria opposifolia ан Daa pteridophylla are a within their respective genera, —
separation of the former as monotypic 1 be discardec
is rejected. Froesia is the REDE CUR ally i most distinguished genus and sister to i
amazonica, once suggested to be recognized a
С once
phic ME +; һа iracte
bus sized.
Lacunaria,
ey Ut
K
‘rnin
ds .
t the generic
guianensis. и is күп most diverse and derived genus with hig
eduction toward smaller нА ‘es and f
with certainty. Proposed close relationship of the tw
al
evel, forms a small but closely related clade with 7:
ghly unresolved relationships and numerous —
flow
wers, fewer stamens, and carpels can be
g the evolution of sex distribution, androdioecy was fixed ea b in a common ancestor of
— and Touroulia, and subsequently dioecy evolved in Lact
‘unaria.
Androdioecy, Froesia, Lacunaria, morphology, phylogenetic analysis, Quiina, Quiinaceae, Touroulia.
8
8
The Quiinaceae are a Neotropical dicotyledonous
family of 51 species, including several taxa not yet
described. It presently comprises Froesia Pires (5
species), Lacunaria Ducke (10 species), Quiina
Aubl. (34 species). and Touroulia Aubl. (2 species).
The Quiinaceae occur principally in primary low-
land rainforests, with a few species found in pre-
—
montane and cloud forests, reaching an elevation
of about 1500 m a.s.l. The family is distributed
from Belize and Jamaica to southern Brazil and Bo-
livia with a center of diversity in the Amazon low-
and forests (Fig. 1), and it comprises shrubs or
medium-sized trees. While the systematic position
of the family within the Malpighiales seems to be
comparatively clear, recent studies have proposed
embedding Quiinaceae in Ochnaceae on the basis
of a single gene, the rbcL (Chase et al., 2000; Sa-
volainen et al., 2000). this systematic
amalgamation may not find support if additional
molecular markers
ered (e.g., Soltis et
However,
and/or morphology аге consid-
al., 2000; Jansen et al., 2001).
The infrafamilial systematics, despite the important
contributions of Pires (1948, 1950, 1953, 1960),
still raises many questions due to our incomplete
knowledge of the family, an issue we wish to ad-
dress in this paper.
Cireumscription and mutual relationships of the
genera of Quiinaceae are issues not hitherto inves-
F, G, GH, IAN . LPB, M, MO, NY,
by dec бош. — he For: vchungsge meinsc cn (Zi 557
from Helge Ax:son Johnsons К und and the
? Botanik/Paliiobotanik,
age 25, 0-6) —
Forschungsinstitut Senckenberg
yartment of Syste
ANN. MISSOURI Bor. GARD.
l We 2 ~ pus ‘tors of * peas s — for the га of mate 2 A. AAU, B, BM, BR, CAY, CEN, COL,
Г, ОСА, R, RB, S, SP, U,
VEN, W, WAG.
The study was supported
/31), “Se ne — — the Naturforschende Gesellschaft, and
» Swedish Natural Science Research Council.
thanke d for their input ar ч constructive criticism оп the manuscript.
Two anonymous re viewers are
and Johann Wolfgang Goethe-Universitüt, Sencke —
Frankfurt am Main, Germany. jschne отт жа uni-frankfurt. ге gzizka@sngkw.uni-frankfurt.c
Jepi matic Botany, Lund University, Östra Vallgatan 18, 22:
Department of Hon Stockholm University, 106 91 Stockholm, Sweden. ulf. а ѕи.ѕе
89: 64—76. 2002.
61 Lund, Sweden. Present address:
Volume 89, Number 1
2002
Schneider et al.
Quiinaceae
EN
د شلک "e
Lacunaria, Quiina
Froesia
Figure |.
species, respectively: Froesia with 5
tigated in a cladistic framework. Based on anatom-
ical data, Froesia was postulated to have a some-
what isolated position in the family (Gottwald &
Parameswaran, 1967). While Froesia forms a ho-
mogeneous, easily identified group, the two Tour-
oulia species differ from each other, especially in
anatomical and the
whether they form a monophyletic group. Delimi-
tation of species and generic boundaries are espe-
cially difficult to discern in Quiina and Lacunarta.
The latter is comprised of a “core group” of similar
species plus a few species more or less aberrant.
The systematic position of Lacunarta oppositifolia
Pires is uncertain, and generic rank for this taxon
has been proposed (Pires, in sched.). Similarly, ge-
neric rank was proposed for Touroulia amazonica
Pires & A. 5. Foster (Pires, in sched.). Quiina pter-
idophylla (Radlk.) Pires is another re spe-
cies, being intermediate between Quiina and Lac-
characters, question arises
unaria in several morphological characters.
In flowering plants, sex distribution is an impor-
tant character for understanding the evolution. of
androdioecy and dioecy. Evolutionary hypotheses
concerning mating systems have been postulated
and reviewed in the light of population. genetics
sex allocation theory) (Ross,
1980; Barrett,
1984;
(sexual selection,
1978, 1982; Bawa,
1981; Charlesworth,
Thomson &
Thomson & Brunet,
Distribution of the neotropic ‘al family Quiinaceae:
species; and Touroulia.
Touroulia
Lacunaria and Quiina, circumscribing 10 and 35
1990: Richards, 1997; Pannell, 1997) or phyloge-
netic 1998).
denced by the existence of androdioecy (and di-
oecy) in different, rather divergent major taxa, il
most lis originated independently and several
times from hermaphroditism (Swensen et al., 1998).
Careful examination shows that most of these spe-
cies are only morphologically androdioecious, but
constraints (Swensen et al., As evi-
м
functionally dioecious because of inviable pollen or
indehiscent anthers. Functional androdioecy is very
rare in seed plants. According to Swensen et al.
(1998) it is known only from four species: Datisca
glomerata (C. Presl) Baill. (Datiscaceae), Mercuri-
alis annua L. (Euphorbiaceae), Saxifraga cernua А.
Gray (Saxifragaceae), and Phillyrea angustifolia L.
(Oleaceae). An additional Schizopepon
bryoniifolius Maxim. (Cucurbitaceae), was reported
by Akimoto et al. (1999). In Quiinaceae, Froesia
has perfect flowers, Touroulia and Quiina are an-
The mor-
species,
drodioecious. and Lacunarta is dioecious.
phologically androdioecious genus Quiina is probably
functionally dioecious too, producing inaperturate pol-
len in morphologically hermaphroditic plants (Pires,
cited after Kubitzki, 1995 pers. comm.; Schneider,
1998).
With respect to the monophyly of Quiinaceae, all
species share a characteristic combination of ana-
tomical characters of wood and bark (Gottwald &
66
Annals of the
Missouri Botanical Garden
Table 1. Characters and character states for the Quiinaceae and the outgroups Elvasia, Ouratea (Ochnaceae), and
Medusagyne (Medusagynaceae).
8. B)
PWNS
an
-l
ہے م
للا МЈ
10.
11.
13.
14.
15.
16.
ae
18.
19.
1
вә
21.
. Stipules free (0); partly fused with deep, subulate lobes (1); completely fused (2).
3.
24.
25.
. Stipule margins entire (0); serrulate (1)
27.
N
N
2:
NN N
© ډ- ټم ف o
ы
—
. Sepals 2 or 3 (0); 4 (1): 5 to 8 (2).
7. Sepal texture membranous (0); coriaceous (1
Plants not caulirosulate (0); caulirosulate (
Terminal internodes terete (0); laterally conspicuously compressed (1).
Leaves alternate (0); opposite (1); verticillate (2).
Adult leaves simple (0); compound (1).
Leaf rachis absent (0); not alate (1); alate (2)
Leaf blade or leaflet ovate (0); elliptic (1); obovate (2
Leaf blade or leaflet glabrous or principal vein pubescent (0); principal vein and leaf surface along the veins
м2
T
pubescent (1); abaxial surface pubescent (2).
Leaf or leaflet apex acuminate to acute (0); obtuse (1); retuse (2).
Leaf or leaflet margin revolute (0); flat (1)
Leaf or leaflet margin entire (0); minutely serrulate (1); serrate (2).
Leaf or pis vein apices not protruding or reaching the margin (0); conspicuously protruding from lamina margin
into teeth (
Number of an less than secondary veins (0); more teeth than secondary veins (1); teeth equal in number to
secondary veins (2); teeth absent (3).
Venation cras} (0); camptodromous (1)
Prominence of leaf venation: abaxially more prominent (0); abaxially and adaxially + equally prominent (1);
adaxially more prominent (i
Intersecondary veins not deve — (О); conspicuously developed (1); present but not conspicuously developed (2).
Tertiary veins scalariform (0); parallel (1); plumose-reticulate (2); reticulate (3
Stomata paracytic (0); anomocytic (1).
Stomata positioned on a flat abaxial surface (0); in small depressions (1).
Petioles canaliculate (0); terete (
Petiole base not conspic ous broadened (0): — broadened, not distinguishable from upper part (1); pul-
vinoid broadened, distinguishable from upper part (2).
Stipules not interpetiolar (due to alternate phyllotaxis) (0); interpetiolar (1); absent (2).
Stipules persistent (О); caducous (1).
Stipules glabrous (0); pubescent (1).
Stipules sessile (0); stipitate (1).
Stipule lobe(s) cuspidate (0); acuminate to acute (1): obtuse (2).
Plants —— (O); andromonoecious (1); androdioecious (2); dioecious (3).
—
—
—
Inflorescence terminal (0); axillar
Inflorescence glabrous (О); inc onspic — pubescent with trichomes mostly —0.2 mm long (1); conspicuously
pubescent (velvety, tomentose) with trichomes mostly >0.3 mm long (:
Inflorescence branched with one main axis (О); brang hed with several axes, fasciculate (1); unbranched (2).
2. Staminate flowers absent (О); in groups of 1 to 3 on terminal branches (1); in groups of 4 to 12 on terminal branches
Pedicel i in upper part slightly widened to cylindrical (0); obconical (
Pedicel articulation between base and 1/3 of the length (0); clearly more than 1/3 of length (1).
Pedicel pud part) glabrous (0); pubescent и (1).
VI
—
. Sepals glabrous (О); margin pubescent (1); margin and abaxial surface pubescent (2).
. Sepals monomorphic, size equal (0); heteromorphic, outer smaller than inner (1).
)
Petal aestivation imbricate (О); contorted (
Petal margin (best observed in bud) glabrous (0); pubescent (1).
Petals obovate (О); elliptical (
Petal apices rounded (О); notched (1)
Functional stamens in hermaphroditic or staminate flowers 10 to 25 (0); 30 to 80 (1); > 100 (2).
Filaments basally free from petals (0); adnate (1).
—
Anthers (in outline) elongate, linear-oblong (0); narrowly elliptic (1); + round (2
Anthe rs opening by apical pores (0); longitudinal slits (1).
. Gynoecium synearpous (0): apocarpous (1).
Carpels 2 (0); 3 (1); 4 to 8 (2); 10 to 14 (3); >16 (4).
Volume 89, Number 1
2002
Schneider et al. 67
Quiinaceae
Table 1l. Continued.
О. Ovules per locule 1 (0): 2 (to 4) (1).
О); all ovules develop to seeds (1).
—
E Some OV ules abort
Fruit a capsule (0); berry (1): follicle
Fruit apex rounded, centrally + concave (0): conical.
Fruits not spotted i with yellowish spots D.
Fruit pericarp in tra
сл ел сл єл єл сл yu
Ve wn
56. Fruits glabrous (О); e nt (
(2); schizoc meric — (3
erse section without lacunae (0): inconspicuous lacunae (1); conspicuous lacunae (2
1).
); dry, indehiscent (4).
e (1).
)
51. Exocarp of mature fruits smooth (0): longitudinally furrowed by underlying lacunae (1): fruit appearing conspicu-
ously ribbed (
58. Seeds к. p pubescent (1).
59. Seeds longer than
60. Crystalliferous cells abundant, in long rows along the veins (О);
wide, ellipsoid (О); as long as wide, globose (1).
—
not abundant, sometimes in groups near the veins
Ol. Crystalliferous cells without special thickening of cell wall (0): always present, with u-shaped thickening (1).
1967).
ous leaf venation with very closely spaced tertiary
Parameswaran, An unusual craspedodrom-
veins is found in Touroulia, Froesia, and Lacunaria
(a slightly different pattern is observed in Quiina),
a set of characters also regarded as unique to the
1950a. b, 1951: Roth, 1996; Zizka
1999). Touroulia
display a characteristic u-shaped
family (Foster,
& Schneider,
Aubl.
thickening of the cell wall (cristarque cells) (un-
Froesia and gui-
anensis
published data). Another common character is the
presence of mucilaginous cavities, particularly in
petioles (Schofield, 1968). and is frequently ob-
served in all genera.
Quiinaceae have formerly been regarded as part
1981). More re-
cently, phylogenetic studies based on DNA se-
quence data (rbcL, atpB, and 185) place the family
in Malpighiales (Fay et al., 1997; Källersjö et al..
1998; Litt € Chase, 1998; Nandi et al., 1998:
Chase, 1996, This position was ac-
cepted in the ordinal classification of the flowering
of the order Theales (Cronquist,
—
pers. comm.).
plants recently published by the Angiosperm Phy-
logeny Group (APG, 1998). These studies suggest
that the closest relatives of Quiinaceae are Och-
naceae and/or Medusagynaceae. Despite this affin-
ity, the relationship between each of these families
is weak. In the study by Nandi et al. (1998), Quiin-
aceae are sister to Ochnaceae when based solely
on weighted rbcL parsimony. When combined with
non-molecular data, Quiinaceae shift and lie basal
to the sister pair of Ochnaceae and Medusagyna-
ceae. The positioning of Medusagynaceae, Ochna-
ceae, and Quiinaceae within the Malpighiales was
unstable depending on character sets used.
For the forthcoming treatment of Quiinaceae for
the Flora Neotropica (Schneider € Zizka, in prep.).
the present study based on morphology attempts to
resolve questions of generic circumscription, in-
cluding morphological synapomorphies. We also in-
vestigate the affinities of the aberrant species Lac-
and
unaria oppositifolia, Quiina pteridophylla,
Touroulia amazonica. Finally, we provide a familial
interpretation of gynoecium morphology and sex
distribution in a phylogenetic perspective.
DATA AND METHODS
TAXA
Ingroup monophyly is necessary for correct root-
ing of the tree (Nixon & Carpenter, 1993). The out-
the
Ochnaceae and Medusa-
group was therefore chosen [топ two most
closely related families,
eynaceae (Medusagyne oppositifolia Baker). Gott-
wald and Parameswaran (1967) already pointed out
that the most closely related family to Quiinaceae
is Ochnaceae, and herein the tribes Elvasieae E nel.
—
and Ourateeae Baill. of the former E
sensu Engler. Hence, two species each of the gen-
era Elvasia DC. Aubl.
Because information on infrageneric relationships
and Ouratea were selected.
within the outgroup is lacking, the species included
in the analysis were selected according to the cri-
teria of (1) their distribution within the geographi-
cal range of Quiinaceae, (2) their previous use in
cladistic analyses (Amaral, 1991), and (3) the num-
ber of specimens available for detailed morpholog-
ical and anatomical studies.
Within the ingroup, a total of 22 species were
Nx as terminal taxa. Since the principal in-
tention of the study was the resolution of generic
relationships within the family, it was considered
redundant to include all species. The present se-
lection comprises representatives of all genera and,
more importantly, all somewhat aberrant species as
well as core representatives from Lacunaria and
Quiina. Froesia is, on one hand, a comparatively
68
Annals of the
Missouri Botanical Garden
homogeneous group; on the other hand, it is the
most distinct genus within the family. Thus, three
species were chosen and regarded as sufficient be-
cause the inclusion of all its species would not
change the principal topology of the cladogram.
Quiina is represented by 12 out of its 34 species,
many of which are difficult to diagnose and circum-
scribe because of remaining taxonomic and nomen-
clatural problems.
CHARACTERS AND CHARACTER STATES
Characters were extracted from morphological
studies undertaken (see Table 1); the final data ma-
trix appears in Table 2. Multistate characters are
generally unordered in phylogenetic reconstruction,
but if there is reason to believe a character state is
intermediate between two other states, such a char-
acter may be ordered (Wilkinson, 1992, 1995). In
our analysis, only two characters (6 and 8) could
be treated as ordered. These refer to leaf outline
and leaf apices where an intermediate transition
state can be perceived. Unknown character states
are coded with a question mark (7) and inapplica-
ble states with a dash (-). Character evolution was
traced by using the software MacClade (Maddison
& Maddison, 1992).
Autapomorphies can either be included or ex-
cluded in phylogenetic reconstruction. Yeates
(1992) argued that removal may also remove infor-
mation from a data matrix, but Bryant (1995) dis-
agreed, saying that they should be removed. Ar-
guments against autapomorphies rest on the fact
that they are unique, uninformative, and inflate the
index (CI).
however,
consistency Phylogenetic reconstruc-
tion, is much more than consistency in-
dices. This study, for example, aims to reconstruct
the major evolutionary lineages within Quiinaceae
with approximately a third of the species of Quiina
sampled. To facilitate future studies using a wider
sampling, autapomorphies were therefore included
to avoid excluding potential synapomorphies.
Polymorphic characters are abundant in Quiin-
aceae and can be treated in different ways. For
example, they could be coded as missing entries,
which then introduce erroneous consistency indices
i 1991).
phic characters could also be included and scored
and tree length (Nixon & Davis, Polymor-
for the observed intraspecific variation, albeit a low
1995). Kornet
Turner (1999) recommended that polymorphic
phylogenetic signal (Wiens, and
characters should be coded as plesiomorphic in fa-
vor of the observed intraspecific variation unless
the ancestral state is unknown. Assessment of the
ancestral state, at least in Quiinaceae, is a critical
point and generally lacking. Since we believe that
polymorphic characters do provide a phylogenetic
signal and resolution, they are scored with the ob-
served states.
Habit.
which means that leaves are crowded at the stem
Plants are caulirosulate in Froesia,
or branch apex (char. 1).
Leaves. The presence of compound versus sim-
ple leaves (char. 4) is problematic with respect to
homology and comparability of character states. For
the current analysis the single leaflets of species
with compound leaves are regarded as homologous
t
equivalent units (Raunkiaer, 1934). Furthermore,
~
» simple leaves because they are functionally
this reflects the equivalent venation pattern of leaf-
lets and simple leaves. Additional support is given
by the presence of simple leaves in seedlings of
pinnate-leaved Froesia venezuelensis Steyerm. & С.
Bunting and pinnatifid leaves in seedlings of
simple-leaved Quiina pteridophylla. In Quiina and
Lacunaria the leaf margin (char. 10) can be incon-
spicuously serrulate, only seen with a strong lens
(magnification > 20X). In this case minute inci-
sions or papillae are seen. The pattern of leaf ve-
nation is a peculiar and particularly important
character in Quiinaceae (Foster, 1950a, b, 1951;
Roth, 1996). All Quiinaceae exhibit a craspedod-
(char. 13:
1979). The presence of conspicuous
romous type
Hickey, 1973,
intersecondary veins (char.
classification following
15) discriminates Qui-
ina and is regarded as a synapomorphy for the ge-
nus. Intersecondary veins have a diameter inter-
mediate between secondary and tertiary veins; they
originate from the primary vein and do not reach
the leaf margins. In camptodromous Medusagyne it
is hard to judge the intermediate veins as distinct
intersecondaries. For the analysis they are consid-
ered as inconspicuously developed intersecondary
veins. The tertiary venation pattern (char. 16) is
unique to the Quiinaceae. In Froesia, Lacunaria,
Touroulia, the tertiary
and veins are densely
spaced, strongly parallel, percurrent or, more fre-
quently, anastomosing at different distances from
their origin (e.g.. Zizka & Schneider, 1999). In Qui-
ina they are less parallel and more conspicuously
branched, anastomosing with intersecondary or ter-
tiary veins, This pattern is called plumose-reticu-
late according to Foster (1950a, b). A scalariform
(ladder-like) pattern is observed in Elvasia.
Stipules. Stipules are always interpetiolar in
Quiinaceae (char. 21), but the number of stipules
or stipular lobes may differ among the genera (char.
22). Stipule number obviously varies with the phyl-
lotaxis. In verticillate Lacunaria there is one stipule
between neighboring petioles (as in verticillate 0.
Schneider et al.
Quiinaceae
Jaquinu JdJOR.IBY")
T OOTTOCOOTO TSOTCOTOO0 театеттоао TOCTOOPTOT OTOOTOOOZT CEOOPCTIOO 'Iqny sisuoupinz Di fmomor
0 OOTTOCOOTO TZOTZOTOOO ELELE OOO TOETOOTTOL OLOO0LOOO0C? q*9000crLO00 SV Y 9S21lq DITUOZDUID DINOMO]
0 OOTTOTOOTO TOOTZTPOOL OPTTPOO0AZ TTCTOOTTOT TTOOCTOOEO 0000TOOTTO еа. тој типп)
0 OOTTOTOTTO TOOTZTPOOT OTATTOOOT? CECLOOLLOL TOTOCTOOEO 0000800110 [ny mdop puim()
0 OOTTOTOOTO TOOTZTOOOT OTTOCTOOTS ea T аео ШЕШ ШШ атообоосоо sang (APLY) Р уЧормә типп,
0 OOTOOTTOTO TAOTZTTIOOL OTCTITOOTT ATCCOOTTOTL TOOOZTOOEO OTOOTOOTTO вәол] Y SAMGg sisuavspd vunne)
0 TOTTTTOTTO TOOTCETOET OECECLOOCC CTCAOTTOOT CLOOCTOOTO ?0*0c00IIO0 sou szsuaoodpio vunne)
0 OPTTOTOOTO TOOTZTPOOT O LEEQUOELL TTCTOOTOOT CLOOCICOIO "0207200190 [AL 2701000 pump)
0 OBYTTOTOOTO TOOTZTPOOT OPTTPOOOTZ ССОО ЕО TOOOZTOOEO 0000100110 [п], »gj&do400ui bum)
0 OOTTOTOTTO TA. DIBIb OD: E OL Ede 0l Tbeb АШ TTOOCTTOCO 0000100110 PUBLI] Y cuoue[q xo nads )]0fi2uo] vunne)
0 OFTTOTOOTO TOOTZTPOOT OTTTTORTTLT PTZTOOPTOL TTOOZTTOTO TTOOADOTTO quy sisuaupmz nunne)
0 TOTOOTTOTO T90ICIOOO0I OTATTEEOCC TICTOPTEOT TODOZTOOLO 0000TOOTTO т, Ppu04 vunne)
0 OTTTOTOOTO TOOTCTPOPT OTTTCOOOTT CLCTOOTTOT TTOOZTOOTO ETOCTOOTTO "qostie) nupidazonao uin)
0 OOTTOTOFETLO TOOTZTOOOT OTTTTOOOTZ TIZTOOTTOT 20002 TTOOTOOTTO "ls `7) "Y DIUOZDUD vunne)
0 TOTOOCTOTO TZOTTOOOBT TIZIATOSTO CEETOOPTOT TLOOT( 0900100190 solid (S2114) prjofiisoddo niDunovT
0 OOTTOCOOTO TEOTZOTOOL OICIIIOOCO CLETOOIICI [9001000€0 000400 “US усу CL) DAYIDISOLIDUL DUDUNIVT
0 OOTTOCOETO TEOTZOPOTE OTETTTRPOZO LEETOOQETE I PEOOTOOOED eooeqoozoo АО) HUB рытрипәр]
0 LOLLOCOOLO ELOTCOTOOO OICIII*0cO LI£IOO?ICI I*"001000cI c00*I00c00 IPN CY[PPY) »40sp2ap nipunov]
0 TOTTOCTOTO LcOLCOLOOOQ OILCIIIOPIO qecroorrct qeoo0roooce c000800200 "US 7) cV (Чї) PIPU Dipunor]
I OPOTOTOOZO TTTTTOCTOO TTCTeTOO00 ADOTOOTOTT 00001000c0 COOOTTITOT 'S `9) Y ишәләр sisuajanzauaa 018201]
E OPOTOTOOZO LIIIIOCIOO LICICLIOO00 q00LOOLOIL OLOOLOOOZ? C*00€*LLIOI зәл] Dd4DJ14] 015904
T OPOTOTORZO TTTTTOCTOO ТТС Те, ОШОО COOTOOTOTT 0000T000Z0 eegQeeTTIOT zonbse 4 "pou Y neam?) psnffip VISIO]
AVAOVNITAO
0002000000 TVOTTOTOOO 0001200020 DU Due С OOOTECTIOO CTLCOTOOTOO 1oxeg piyofipsoddo au{@Dsnpa J
AVAIVNADVSAGAN
0 00000000€I 0c00000000 1009200000 09010000-0 0000711000 5000900000 Te Coq) »aograapd pompano
T 00000000€T 0200000000 т002с00000 0Р0Тт0000-0 POOOCTIOLL c000100000 Puy (Puny) $US] Danan)
0 0600000070 0000000090 0000P00000 0*000000-0 00000000Е0 0000100000 SI) ('Qoue[q) saproispa]a DISDA]H
$ 0600000070 0c00000090 0000P00000 09000000-0 *?00^000000 1000100000 "GOV pajpdoppo DISD]
AVADVNHDO
O68L9STECT O68L9SPECT 068L9STtCI 068L9STtcCl O68L9STECL 068L9STvtcl
9999959955 SyvVYVVVYVY VEECECEEEE COGCCECGEGG GLELECETEL T
Volume 89, Number 1
2002
‘BPP SUISSIWL =
"xugeui əy} UT ‘(QRAORUABRSNPa|A ) au&Spsnpapy — (әвәәвицә()) DANA *mispa]y *dnougjno
E/E = 90/10 =
P ‘2/0 = 2 7
“1/0 = F S199] Ҷим exe] orydaousjod pue (-) usep e цим papoo әле s9]e]s a[qeorddeur
əy} pue ara2eurın() dnoasur əy} Jo siojoe1euo [eorgo[oud.our [9 jo xuqeu eC "c APL
70
Annals of the
Missouri Botanical Garden
pteridophylla). In opposite-leaved Quiina and Lac-
unaria oppositifolia there are four per node (paired
between petioles), while in opposite-leaved Tour-
oulia only two per node are observed. In the latter,
this apparently single stipule is interpreted as a
product of paired fusion similar to cases observed
in Rubiaceae (Goebel, 1932). In opposite-leaved
Froesia, the two stipules per node are deeply di-
vided into setose lobes. In Quiinaceae, stipules are
present on the terminal node at least. If stipules
are generally lacking on the more basal nodes they
are coded as caducous (char. 23); if generally pre-
sent on the three uppermost nodes they are treated
Sex distribution (char. 28) is hetero-
geneous in Quiinaceae, with Froesia being bisex-
ual, Lacunaria unisexual and dioecious, and Qui-
ina and Touroulia morphologically androdioecious.
with male and hermaphroditic flowers on different
plants. The number of sepals (char. 36) is variable
in Quiinaceae. Four sepals are common and usually
constant for a species, while species with five or
more sepals display a higher variability in number.
Less than four sepals is an exceptional condition
and therefore given its own state.
The hermaphroditic Froesia flower bears numer-
ous stamens, frequently more than LOO (char. 44).
In the androdioecious genera, the staminate flowers
generally produce 30 to 80 stamens, whereas the
hermaphroditic flowers normally produce fewer. For
coding, only the flowers with functional stamens are
included in the analysis. Character 45, filaments
free or adnate to the petals, only refers to the her-
maphroditic flowers of species of Quiina because
in that genus staminate flowers do not show this
48), Froesia
is clearly apocarpous while Ouratea is syncarpous,
trait. Concerning the gynoecium (char.
because apocarpy is only gained during fruit de-
velopment (Amaral, 1991). In Lacunaria and Tour-
49) is variable
and relatively high (4 to 14), while in Quiina there
oulia the number of carpels (char.
are commonly 2 carpels. In the latter, only a few
species (e.g.. Quiina florida Tul., Q. paraensis Pires
& Fróes) show transitional states with up to 5 car-
The exocarp is usually longitudinally
striate to furrowed by underlying resiniferous la-
cunae that are more or less conspicuous in tran-
section (char. 57). In Quiina florida and Q. paraen-
sis, the exocarp appears quite smooth and is not
striated by the lacunae. Additionally, their fruits are
spotted by a substance that looks like dried resin-
ous exudate (char. 54).
Crystalliferous cells. One characteristic feature
is the presence of the cristarque cells (char. 61), a
term introduced by van Tieghem (1902). These are
crystal-bearing cells with a conspicuous u-shaped
wall thickening. They are often cited in familial
descriptions (Cronquist, 1981, as solitary crystals;
Amaral, 1991; Bhattacharyya & Johri, 1998; Ku-
bitzki, 1995 pers. comm.) and hence could be er-
roneously assumed to be characteristic for the en-
tire family. So far we were able to confirm them
only for the genera Froesia (F. tricarpa Pires, E
venezuelensis) and Touroulia (T. guianesis), in ac-
cordance with the findings of Foster (1950b). In
addition to these specialized cells, crystal druses
can be regularly observed in the leaf tissue (char.
60).
PHYLOGENETIC ANALYSIS
r €
The data matrix in Table 2, containing 27 taxa
and 61 characters, was analyzed with PAUP* 4.0
for Macintosh (Swofford, 1998), using the branch
and bound algorithm under the assumption of Fitch
parsimony (Fitch, 1971). Multiple character states
were interpreted as uncertain. An initial search was
undertaken with equal weights saving all optimal
trees, To evaluate characters with the strongest phy-
logenetic signal and to choose among equally par-
1988, 4). succes-
1969)
The settings
simonious trees. (Carpenter,
sive weighting analysis (Farris, was
undertaken after the initial search.
used the rescaled consistency or RC index (Farris,
1989),
set to 1000. The process was reiterated until the
A similar
and to avoid fractions the base weight was
—
same tree length was obtained twice.
analysis was carried out where characters 6 and 8
were ordered.
Branch stability was estimated with Bremer sup-
port and jackknife analysis for both the equally
weighted and weighted characters. Bremer support
is defined as the number of extra steps necessary
› lose a clade in the consensus tree, using the
converse constraints approach (ат
1994; Kiillersjó et al., 1992; Farris, 1996). it
weighted and rescaled branch support values cal-
culate the robustness for each branch in the weight-
1994; Gustafsson &
To ease the construction of all nec-
ed consensus tree (Bremer,
1995).
essary constraints, the computer program Auto-
Each con-
Bremer,
Decay 4.0 was used (Eriksson, 1°
straint was estimated with a heuristic search of 100
replicates of random additions of the taxa (10 rep-
etitions of each replicate), tree bisection-reconnec-
tion (TBR) branch swapping, holding five trees at
each step, and saving all equally parsimonious
trees. Jackknife (Farris & al., 1996) investigates the
structure, or phylogenetic signal, in a matrix with-
Volume 89, Number 1
Schneider et al. 71
Quiinaceae
Felsenstein,
—
out permutation, contrary to bootstrap
1985). but excludes an assigned fraction of char-
acters, here set to 35%. The search strategy was
set as for the Bremer support, but with 1000 rep-
licates and saving no more than 100 trees.
RESULTS
The analyses using morphology to infer the phy-
logeny of Quiinaceae yielded 1643 most-parsimo-
nious trees using all characters as unordered as
well as when characters 6 and 8 were ordered. The
trees are 163 steps long (when ordered 164 steps)
with an RI of 0.775 and a CI of 0.561, or 0.529
when uninformative characters were excluded (Fig.
2). Counting steps within the polymorphic taxa, the
trees are 282 steps long, indicating numerous poly-
morphic characters.
Successive weighting of the characters yielded a
stable result of 2 trees after three iterations, 73.206
steps long using unordered characters. This tree is
a subset of the initial 1643, resulting in a consen-
sus with a much better resolution of the otherwise
collapsed genera Lacunaria and Quiina. Consensus
of the primary 1643 trees and the 2 trees obtained
after successive weighting together with jackknife
fractions, Bremer support values, and weighted and
rescaled support values, are shown in Figure 2.
One of the most-parsimonious trees with characters
optimized on the branches is shown in Figure 3.
Referring to the unweighted analysis, Quiinaceae
form a well-supported monophyletic group, a grade
from Froesia at the base to the most derived genus
Quiina. Within the family, all genera are monophy-
letic except for Lacunaria, a genus completely col-
lapsed to a comb. Based on equally weighted char-
acters, no resolution within Quiina can be
retrieved. The small genus Touroulia forms the sis-
ter to the Lacunaria—Quiina complex. Support for
Froesia and Quiina must be regarded as strong,
while support for Touroulia is moderate. No clear
signal for a monophyletic Lacunaria can be re-
trieved, and L. oppositifolia attaches as sister to
Quiina, a position with low jackknife support of
54%. but not found in the consensus (Fig. 2).
As to the weighted analysis, resolution is im-
proved and only one trichotomy remains, the one
in Froesia. All clades found in the equally weighted
analysis are retrieved and, as a general trend, mod-
erately or strongly supported groups are often better
supported after successive weighting. Froesia, sup-
ported by a maximum jackknife value in the equal-
ly weighted analysis, needs five extra steps to be
lost when unweighted, a value twice as strong after
successive weighting. A similar situation is ob-
served for Touroulia, and the support for Quiina
increases almost three times. Using the weighting
approach, Lacunaria is now found to be monophy-
letic, although with a Bremer support of only 0.9.
DISCUSSION
Because a cladistic analysis can only elucidate
ingroup relationships, we cannot put forward a hy-
pothesis of relationships within the order Malpigh-
iales, or determine whether Quiinaceae are sister
r sister
to a paired Medusagynaceae—Ochnaceae, «
to either one separately. Support for Quiinaceae 15
strong, with little difference in support values be-
tween equally weighted and weighted characters.
This implies that several characters are initially
strong and basal on the tree; these non-homopla-
sious synapomorphies include pubescent stipules
(char, 24
eromorphic sepals (char. 39: 1), and a fruit exocarp
with lacunae (char. 55: 1; Fig. 3). Other synapo-
morphies for Quiinaceae, addressed in the intro-
1), pubescent sepals (char. 38: 2). het-
duction, are the unique leaf venation pattern with
densely spaced parallel or plumose-reticulate ter-
tiary veins (char. 16) and the interpetiolar stipules
char. 21: 1).
In a molecular study of the relationships of Me-
dusagyne oppositifolia by Fay et al. (1997), Quiin-
aceae were represented by Quiina pteridophylla,
—
Touroulia guianensis, and Lacunaria jenmanit
(Oliv.) Ducke, all included in our study. Contrary
to our results, Quiina was found to be the most
basal taxon in the family. Quiina pteridophylla is a
morphologically aberrant species within the genus
with some characters resembling species of Lacu-
naria, such as the phyllotaxis and the stipule num-
ber. Therefore it may be an inappropriate represen-
tative of Quiina and could attach as sister to all
other genera due to long-branch attraction.
Froesia is indicated by our cladistic analysis to
be the most basal genus in Quiinaceae, followed by
a grade of Touroulia, Lacunaria, and Quiina. Par-
ticularly noteworthy is the apocarpous gynoecium
(char. 48) of Froesia, a character state known nei-
ther from other Quiinaceae nor from Ochnaceae or
Medusagynaceae. However, some Ochnaceae ex-
hibit a secondary—otherwise called “ecological”
Baum, 1951)
(Amaral, 1991). In her cladistic analysis, Amaral
(1991) interpreted this secondary apocarpy as a de-
rived state differing from the states observed in the
Quiinaceae and Scytopetalaceae (both families
apocarpy during fruit development
—
therein used as outgroups). Whether apocarpy in
Froesia evolved secondarily is difficult to infer. The
existence of a compitum, providing support for the
72
Annals of the
Missouri Botanical Garden
Medusagyne oppositio.
Elvasia calophyllea
Elvasia elvasioides
Ouratea lucens
Ouratea parviflora
Froesia diffusa
апо то -
Froesia tricarpa
100/100
5/10.8
[100/100
mM
— Touroulia amazonica
Froesia venezuelensis
69/87
4/5.5 1
12.6
~ = Touroulia guianensis
-161 |
Lacunaria oppositifolia
Lacunaria macrostachya
– 10.9
78/88
2/3.5
Lacunaria jenmanii
Lacunaria crenata
Lacunaria decastyla
Quiina paraensis
Quiina macrophylla
[ — Quiina longifolia
—— Quiina rh ytidopus
Quiina florida
Quiina tinifolia
г Quiina cruegeriana
L— Quiina guianensis
Figure 2.
morphologic ‘al characters (Tables | ane 2),
analyses yielded 1643 trees (163 s
other, Dotted — indicate coll; — Г anc ches i in “the prime
weights (left) a
ranches are Bret mer support values, i.
and weighted and rescaled characters (right)
view that apocarpy is derived (see Endress, 1982;
Kubitzki, 1995 pers. comm.), could not be ob-
served. Nevertheless, according to the present anal-
ysis, this character state is considered a synapo-
morphy for Froesia.
The cladistic analysis indicates an isolated po-
sition for Froesia in terms of morphology. Support
well as weighted — ‘ters (right) from the primary ind
.. additional steps needed to c dines. a node for unweighted characters (left).
Quiina amazonica
Quiina obovata
Quiina oiapocensis
Quiina pteridophylla
Strict consensus tree of Quiinaceae retrieved from a Волаи cladistic analysis applying к qual we ights to
econdary analysis using s successive The
= two most-parsimonious the
гу analysis — a ше fractions above 50% using equal
are snc
for the genus is strong with no less than eight syn-
apomorphies (Fig. 3). These results correspond to
the findings of Gottwald and Parameswaran (1967),
who even proposed, based on anatomical studies,
the establishment of a separate subfamily for Froe-
sia. Besides the apocarpous gynoecium, deeply di-
vided stipules with setaceous lobes (in all but one
73
Volume 89, Number 1 Schneider et al.
02 Quiinaceae
Ф
зай
5 qc S
© £ y
DES TS
SG Ф Ф © o Q
D EE AS оо о ®
кх у jk yO SP лт о ол д
сєєк CG CG CG О
90 2:0
0
17:1 — SE
38:2 14:1
as 44:1 „22:2
Td 531 ma 26:1
56:1 4-370
9.0 E691
14:2 25:1
31:1 351
33 36:2
K19:0 6:2
ا — 23:0
20:2
1:1 9:1
5:1 10:1
22:1 12:1
+ 40:1
ЩИ 43:1
"2442
щи 48:1
m" 49:1
ща 522
ааз
30:1
"m 283 312
35:0
07 38:1
49:0
21
15:1
со 4:1 16:2
12:2 19:0
16:1 40:0
45:1
R 81
30:2
а, 16:0
= 12:3
pos 20:1
: 29:1
85:1 CJ 36:1
57:1 CH 41:1
Ed Legend
23:1
on ши no homoplasy
46:2
52:1
55:2
58:1
>< reversals
f the two most-parsimonious trees from the successive weighting analysis of the Quiinaceae with
Figure 3. One o
characters optimized on the branches (number left of colon = character: right = character state, cf. Table 1).
groups (char. 32: 1). a berry-like fruit (char. 52: 1),
and pubescent seeds (char. 58: 1). Touroulia forms
a monophyletic group and is recognized by only one
synapomorphy, the compound leaves with an alate
species), glabrous seeds, and hermaphroditic flow-
2) As previously mentioned, ge-
ers are important characters to distinguish the ge-
nus. This example addresses the importance of in-
cluding autapomorphies in phylogenetic. analyses.
following Yeates (1992). If only one Froesia species rachis (char. 5:
nerie rank was proposed for T. amazonica (Pires, in
sched.). a view also advocated by Gottwald and
Parameswaran (1967) based on their anatomical
had been included, a number of characters would
studies. Nevertheless, from a morphological point
have been regarded as autapomorphies. uninfor-
of view, the two Touroulia species are very similar
mative, excluded, and hence probably overlooked
as potential synapomorphic characters for the ge-
nus. \
Apart from Froesia, the remaining species of (Zizka & Schneider, 1999), confirmed by the pre-
Quiinaceae are distinguished by several synapo- sent analysis, and, hence, Pires’s proposal is re-
morphies, including the evolution of staminate flow- jected.
A number of characters unite the genera Quiina
ers (char. 28: 2). their position in small terminal
74
Annals of the
Missouri Botanical Garden
and Lacunaria. Polymorphic characters such as the
position of the inflorescence (char. 29) in Lacunaria
are interpreted as inconclusive of relationship (soft
1999). Regarding
Lacunaria, conflicting data fail to conclusively re-
starts sensu Kornet and Turner,
solve the taxa in the analysis using equal weights,
but support for Quiina is strong. However, a phy-
logenetic signal for Lacunaria is picked up if the
approach to character weighting is used, gaining a
fully resolved tree, but with very low support val-
ues. A jackknife value of 61% and a rescaled Bre-
mer support value of 0.9 cannot be considered as
much support for a monophyletic genus. Based on
the weighted analysis, a possible single synapo-
—
morphy is identified for Lacunaria (char. 28: 3).
Most other characters are homoplastic within the
genus, and some are also observed in, for example,
Quiina pteridophylla. Our knowledge of interrela-
tionships within Lacunaria is not satisfactory, but
pending future analysis we refrain here from pro-
posing any systematic rearrangement and treat the
genus in its traditional circumscription,
3: 2)
and completely fused stipules (char. 22: 2), have
Two characters, verticillate leaves (char.
evolved in the clades of Lacunaria and Quiina.
Based on our morphological studies, Оила pteri-
dophylla was expected to be closely related to Lac-
unaria. The present analysis indicates that Q. pter-
idophylla shares almost all other generic characters
and that it is deeply nested within Quiina rather
than having a basal position in the genus. Forcing
a sister relationship of Q. pteridophylla with L. op-
positifolia, which it superficially resembles, costs
no less than 10 extra steps. Thus, there is no doubt
that these two species are not closely related and
cannot form a sister relationship.
Reasons for previous hypotheses of recognizing
Lacunaria oppositifolia at generic rank can here be
understood. Following the weighted analysis and
D
tracing. character. evolution. (Fig. 3), the species
forms a clearly distinct branch with no less than
five homoplastic characters. The remaining lineage
of Lacunaria has, interpreted from the tree in Fig-
ure 3, gained the same number of homoplastic
characters, although they are different. Although a
there 1
Л
monophyletic genus сап be envisioned,
also weak jackknife support for L. oppositifolia
forming a separate lineage nested between Lacu-
naria s. str. and Quiina if equal weights are used.
One option is to include £. oppositifolia in Quiina,
but this solution is not attractive because a mor-
phologically well circumscribed group would be
lost. Leaf venation pattern, the absence of interse-
condary veins, and the number of carpels do not
coincide with the generic concept of Quiina. More-
over, cladistic support of the genus Quiina is almost
four times as strong as the next internal node. A
second solution is to describe a monotypic taxon,
but due to the weak support, and not appearing in
the consensus, the issue of the rank and position
of L. oppositifolia has to be postponed until other
data, particularly from DNA analyses, are obtained.
Taxa sampled of Quiina for this analysis include
about one third of the total number of the species.
As a genus, Quiina is rather well supported, but
support clearly increases after successive weight-
ing. A cursory inspection of Figure 3 also reveals
a number of reversals and parallelisms on branches
leading to the terminals, especially in the most ad-
vanced part of the tree. We interpret this, in con-
junction with a high ratio of polymorphic charac-
ters, to indicate that novelties are allowed to rise
and spread within the lineage. They represent soft
reversals without hard starts of character states
(Kornet & Turner, 1999). Two reasons, and possibly
more, explain this observation. First, the lineage
might be fairly young and presently going through
numerous speciation events without character state
fixation. Second, interpretations of our character
states may be incorrect, i.e., misinterpretation of
homologies.
EVOLUTION OF ANDRODIOECY
Different modes of sex distribution in the Quiin-
aceae—androdioecy in Quiina and Touroulia, di-
oecy in Lacunaria—seem to have evolved from a
hermaphroditic ancestor. Hermaphroditism is the
plesiomorphic state, found in the basal genus Froe-
sia and in the outgroup Ochnaceae. Andromonoecy
is an autapomorphy for the outgroup taxon Medus-
agyne and is not further considered herein. Andro-
28: 2) is :
Quiinaceae that evolved,
dioecy (char. synapomorphy within
m-— "*
-
—
phylogenetic per-
spective, in the following pathway: androdioecy is
the plesiomorphic state and dioecy independently
evolved in Lacunaria.
Concerning the evolution of androdioecy and di-
oecy, several models have been formulated (Ross,
1978, 1982; 1981; Charles-
worth, 1984; 1990; Richards,
1997; Swensen et al., 1998). The majority regards
dioecy as derived from hermaphroditism via andro-
Thomson & Barrett,
Thomson & Brunet,
dioecy or gynodioecy (Darwin, 1877; Westergaard,
1958; Вама. 1980: Richards, 1997). Another model
1982)
is from hermaphroditic flowers via andromonoecy,
of androdioecy evolution proposed by Ross (
as observed in some Solanum species. However,
andromonoecy is only known from the outgroup.
Medusagyne. In contrast, Swensen et al. (1998) ar-
Volume 89, Number 1
2002
Schneider et al. 75
Quiinaceae
gued that [functional] androdioecy evolved from di-
оесу in Datiscaceae. This path is improbable for
Quiinaceae since androdioecy evolved once in the
ancestor of Touroulia, Lacunaria, and Quiina, with
a further reduction to dioecious flowers in Lacu-
пага
Mechanisms that favor the evolution of andro-
dioecy and dioeey are explained by models of sex-
ual selection and sex allocation, and are discussed
by several authors (Ross, 1978, 1982: Bawa, 1980:
Barrett, 1981; Charlesworth, 1984:
1990; Richards, 1997; Pannell,
1997). One hypothesis argues that in female-sterile
Thomson &
Thomson & Brunet,
plants the loss of seed production is more than
compensated for by the reallocation of resources to
increased pollen production (Thomson & Brunet.
1990). This may be achieved through an increased
number of staminate flowers or stamens, or through
larger anthers. Flower numbers are more or less
equal among the Quiinaceae or even lower in the
dioecious taxa.
—
(andro- Anthers are roughly of the
same size in the family, hence not serving as a use-
ful argument. The number of stamens, indeed,
generally increased in staminate flowers of Quiina
and Touroulia. Nevertheless, comparing the genera,
the stamen numbers do not provide an explanation
for the evolution of the observed pattern of sex dis-
tribution since the supposedly ancestral hermaph-
roditic flowers of Froesia bear the highest number
of stamens within the family. If we trace an evolu-
tionary pathway from hermaphroditism through an-
drodioecy to dioecy, as postulated above, we see a
reduction of stamen numbers from more than. 100
in Froesia, ЗО to 80 in Touroulia, most Lacunaria,
and some of the Quiina species, to approximately
10 to 25 in Quiina and in a few Lacunaria species.
including £L. oppositifolia. Thus, other mechanisms
such as the avoidance of inbreeding may provide a
better explanation for the evolution of dioecy in the
present case.
Literature. Cited
Г. Fukuhara < K. Kikuzawa.
variatii
chizopepon bryoniaefolius (Cucurbitaceae).
Amer. J. Bot. 86: — ).
Amaral, M. C. К. 1991. Phylogenetise he — matik der
Ochnaceae. Bot. Ji ene Syst. 113: 105-190.
APG. 1998. An ordinal classification for
flowering plants. Ann. Missouri Bot.
253.
Baum, H. 1951. Die Frucht von Ochna — ein Fall
ökologischer Apokat
„= К. 5. 1980. Evolution of diseno d in c ants.
. Rev. Ecol. Syst.
Blas tharyya, B. & B.
Akimoto, J..
and ¿ genetic
1999. Sex ratios
on in a functionally androdioecious
— s of
Gard.
тре. Oesterr. Bot. Z. 383-39.
15-39,
M. Johri. 1998. Flowering Plants:
Taxonomy and Phylogeny. р r- Verlag. Naruso Pub-
9
lishing Hones New
Bremer, К. 1988
in angiospe rm phyloge netic reconstruction. Evolution
12: 795-803.
12 Se a h support and tree stability. Cladis-
lies 10:
Bryant, Н. С po Why ol uelis — 2t re-
moved: A reply “ ha o Cladistics 11: 381—
Carpenter. J. M.
. The limits of amino acid sequence data
Ж ES: among "ийн ET
peu С Ang € adis stic s4:
— ———,. 1994. Successive weighing re UE pa evi-
den Сыйга 10: 215—220.
Androdioces y and the evolution of
. ya к 23: 333—348.
Fay
Chase „М ау & У. Savolainen. 2000. Higher-
level classification. in ka angiosperms: New insights
from the perspective of DNA sequence data. Taxon 19:
Cronquist, А. 1981. An Integrated System of Classification
J
£
of Flowering z P ш Columbia Univ. Press.
. Different E Fonos of Flowers on P n ШЕ vl
the Same Spe cies. Murray, Londor
К. 1982. -» arpy and ils alternative modes
or
Darwin. C.
Endress. P.
of escaping disadvantages of apocarpy in primitive an-
Taxon : 8—52.
1998. AutoDecay, version 4.0%. Stockholm.
"n.
Farris, J. S. 1969. A successive ии approac h
to character weighting. Syst. Zool. 4-385.
. 1989. The re tention” index бе m rescaled con-
sistency ‘ude x. Cladistics 5: 417—419
1996. Names and origins. Cladiaiños 12: 263—
2
—, V. A. Albert, M. Kallersjo, D. Lipscomb & А.
Kluge. 1000. Parsimol tgs — outperforms ne "i
—124.
bor- joining: Cladistics 11:
\
Fav. M. ‚ Swensen ^ 3 W. Chase. 1997. Taxo-
nomic aliinitie s of d uis oppositifolia (Me dusa-
evnaceae). Kew Bı S : 111-120.
Felsenstein, J. 1985. E e limits on phylogenies:
An approach using the bootstrap. Evolution 39: 783
91.
Fite h, W. M. 1971.
lution: rue —
106-416.
Towards defining the course of evo-
ora seoil tree topology.
ysl. Zool.
ALS. T n
in 1 Quiina ae n Ducke.
— and venation г tha leaf
Amer. J. Bot. 37: 159-
Foster,
1950b.
Touroulia guianensis Aubl.
\ Bot. 37
Ve — and histology of the leaflets in
and Froesia tricarpa Pires.
mel Яир M. > —
1951.
hes
Heterophyly and foliar venation in Lacu-
— Bot. Club 78: 382—400.
gmap der Pflanzen. Part 3, 3rd
naria.
Goe be
ed. — pri
Gottwald, H. & N. Par
atomie und Systematik der dent d eae. Bot. Jahrb. Svst.
81.
H. G. « K. Bremer. 1995. Киро ову апа
ıylogenelic interre els of the Asteraceae, C:
lyceraceae, Campanulaceae, Goodeniaceae, andı re late
rales). Amer. J. Bot. 82: 250-205.
Mickey, L. 1973. Classification. of pr architecture. of
to ns leaves. Amer. J. Bot. 60: 17—33.
A . A revised classification P E architecture
of dos EMI leaves. Pp. 25-39 in C. R. Metcalfe
erlag. Jena.
rameswaran. . Beitrüge zur An-
)
) O
TNR M.
—
— (Aste
76
Annals of the
Missouri Botanical Garden
& L. Chalk (editors),
Clarendon Press, Oxford.
Jansen, S., P. Baas & E. Smets.
occurrence and systematic importance in eudicots. Tax-
Anatomy of Dicotyledons
‚ Vol. 1
on 50: 135-167.
Kiillersjé, Farris, A. G. Kluge & C. Bull. 1992.
15-2
Skewness and pe е Cladistics 8: 27!
——, M. Chase, B. Bremer, M. oF
Humphries, G. Petersen, O. Seberg & K. Bremer. 1998,
Simultaneous parsimony jackknife analysis of 2538
rbcL DN! reve M support for major clades
of green gem land plants, seed plants and flowering
plants. PI. — 213: 259-287.
— D. Turner. 1999, Coding ee
— — reconstruction. Syst. ‘Biol. 48: 365
sequences
79,
Litt, А. & М. W. Chase. 1998. The systematic position of
Fuphronia, with comments on the position of e
An m based on rbcL sequence data. Syst. Bo
01—40
Maddison, y P & D. К. ا
Analysis of Phylogeny Chara
3.0. Sinauer, Sunderland, Massac
Nandi, O. L, M. W. Chase 8 P. K Endres. 1998. A
combined cladistic analysis of angiosperms using rbcL
and non-molecular data sets. Ann. Missouri Bot.
85: 137-212
Nixon, K. C. & J. M. ا 1993. On outgroups. Cla-
distics 9: 413-420
& Davis. laxa, ‚ missing
values and e lacis 7: 233-241.
Pannell, J. R. . The maintenance of la and
a in a me tapopulation. E volution 51: 10-20.
Pires, J. M. 1948. Quiinaceae. Pp. 22-32 in G. A. Black
& J. M. Pires (editors), Deia géneros novos Curupira e
Froesia, cinco espécies novas, uma nova combinag йо,
1992. MacClade:
r E SIMI. vers.
Gard.
1991, фо
* analysis. Cladist
с — e observações sobre ги is
. Bol. Técn.
ed região amazón-
Inst. Agron. N. 15: 1-32
950. Contribuia io Ares a
Quiinaceae, Bol. Téen. Ins
19
isis Amazónica.
. Agron, N. 20: 41—52
2.
lanis novas i Amazónia. Quiinaceae.
Bol. Teen, * Agron. М. 28. 45—17.
. 1960. Plantas novas da Amazonia, Quiinaceae.
Bol. Técn. Inst. — N. 38: 23—
Raunkiaer, C. 1 The Life Forms of Plants and Statis-
tical Plant Анан C isl ndon Press, Oxford.
Richards, A. J. 1997. Plar маны Site ms,
Chapman and S London
Ross, M. D. 19 The evolution of gynodioecy and sub-
а Pus 32: 174-188.
2nd ed.
2001. Vestured pits: Their
. 1982. Five evolutionary иши to subdioecy.
Amer. ote 119: 297-318
Roth, I. 1996. Microscopic venation patterns of leaves and
ч ir ните іп the distinction of (tropical) species.
T -19 — al. (editors), Encyc —
{ан Anatomy Vol. 4. Gebr. Borntraeger, Ber
lan . E Fay, г a Albach, A. Backlund, м.
van der Ban > M. Cameron, S. A. Johnson, M.
Lledó, J.-C. Pintaud, M. Powell, M. _ Sheahan, D. E.
Soltis, P. S. e Р. Weston, W Whitten, K. J.
Wurdack & M. W. Chase. 2000. ias ny of the eu-
dicots: A dic — le on — based on
rbcL gene se Ve * es. Kew Bull. 257-309
Se acta ar J. V. 3. El género Quiina (Quiinaceae) con
especial referenc ia a las especies de Venezuela. Acta
Bot. 1—
Schofield,
—
—
А 1968, Petiole anatomy of the —
and re — yd Mem. New York. Bot. Gard.
—55.
Soltis, D. E., P. S. Soltis, M. W. Chase, M. E. Mort, D. C.
Albach, M. Zanis, V. а ОЕ n, "X H. Hahn, S. E
Swensen, L. M. Prince,
. Farris. 2000. Angio-
— phylogeny infe med from 128 aper rbcL, and
tpB se a nces. js ot. J. . Soc. 133: 461.
Swe nsen, 5. M., a «T bis & uas 8 rg. 1998.
site Monophyly and e se ey nce of
breeding s vem evolution, Syst. Bot. 23: 157-169,
Swofford, D. L. . PAUP*, Fine p Us-
ing тее — other methods). У
Sunderland, Massac jor s.
Thomson, J. & 5 H. Barrett. 1981. Selection for
oute rossing, se xual di and i evolution of
oecy in plants. Amer. Naturalist 118: 443—449.
—— & J. Brunet. 1990. Hy;
f dioecy in — plants. TREE
Kress,
tig М.К. Кау, М. de 1, S
W. J 1. Nix
Datiscaceae rev
. 4. Sinauer,
` di-
vhs for the evolution
: 11-10.
Tie gu чп, P ~ 102. Sur les Oc ino ées. Ann. Sci. Nat.
Bot., de l —
Weste — 195 Тһе mechanism of sex determi-
nation in s ‘ious owe ‘ring plants. Advances Genet.
9: 217-28
Wiens, J. J. 1995 5. Polymorphic characters in phylogenetic
systematics. Syst. Biol. 44: 482-500.
W — M. 1902. Ordered versus unordered charac-
ane stics 8; 375-385.
‚А comparison of two methods of character
— ов. Cladistics 11: 297-308.
Yeates, D. 1992.
: 387-389,
Zizka, G. & J. V. Schneider. 1999, 5
Aubl. (Quiinaceae). Willdenowia 29:
Why remove autapomorphies? Cladistics
genus Touroulia
A REVIEW OF THE GENERA Warren D. Hauk’
ROENTGENIA AND
POTAMOGANOS
(BIGNONIACEAE)!'?
ABSTRACT
Roentgenia and Potamoganos (Bignoniaceae) are small genera of lianas with “Cydista-type” corollas that are closely
related to larger and n ore well known genera such as C — ‚ Clytostoma, and perhaps Phryganocydia. Roentgenta is
ditypic, containing R. bracteomana and " sonda whereas Por tamoganos is monotypic, containing only P. microcalyx.
The two species of Roe nigenia have white to lavender, campanulate-funnelform, Cydista-type e stems with eight
phloem arms, bi-trifid tendrils, 3(—4)- ме. ollen with reticulate exine, 2-seriate ovule organization, nectariferous
disk lacking, linear-oblong fruit, and winged : um Roentgenia a is distributed from southern Colombia lo
northern Bolivia and north-central B scil. whereas A. sordida is restricted to eastern. Venezuela, Guyana, Suriname,
French Guyana, and portions of northern Brazil. Potamoganos нир, ias lavender to magenta, testis
funnelform, Cydista-type corollas, with four phloem arms, trifid tendrils, 3-colpate pollen with retic wikis xine, d-seriate
ovule organization, and a nectariferous disk. Potamoganos microcalyx is known from southern Venezuela, Guyana,
Suriname, and northernmost Brazil. A key to species, maps of species distributions, reproductive phenology. and
illustrations of R. bracteomana and p microcalyx are provided.
Key words: Bignoniaceae, Bignonieae, Cydista, Potamoganos, Roentgenia.
Roentgenia (K. Schum. ex Sprague) Urb. and Po- distinguished by a combination of characters: stems
tamoganos Sandw. (Bignoniaceae) belong to tribe with four phloem arms, trifid tendrils, 3-colpate
Bignonieae, which is composed almost exclusively pollen with reticulate exine, 4-seriate ovule orga-
of lianas and which has been demonstrated recently nization, and a nectariferous disk (Gentry, 1977,
to be monophyletic using rbcL and ndhF gene se- 1978: Gentry & Tomb, 1979; Tomb & Gentry, un-
quences (Spangler & Olmstead, 1999). Roentgenia published; Hauk, 1997). Potamoganos has a “Cy-
is a ditypic genus of primarily lowland lianas with dista-type” corolla (Gentry, 1978) and is closely re-
white to lavender, campanulate-funnelform, “Cydis- lated to several larger genera (Table 1) such as
ta-type” corollas (Fig. 1), stems with eight phloem Cydista, Clytostoma, Roentgenta. p — y-
arms, bi-trifid tendrils, 3(—4)-colpate pollen with dia (Gentry & Tomb, 1979; Hauk,
reticulate exine, 2-seriate ovule organization, nec- The late Alwyn Н. Gentry — a
tariferous disk lacking, linear-oblong fruit, and and/or Potamoganos in several regional floras
winged seeds (Gentry, 1977, 1978, 1997; Gentry & (Gentry. 1977, 1978, 1983, 1993, 1997. in press).
Tomb, 1979; Tomb & Gentry, unpublished). Roent- but his monographie work did not encompass
вета is closely related to such genera (Table 1) as Roentgenia, Potamoganos, or other genera of the
Cydista Miers (6 species), Clytostoma Miers ex Bu- tribe Bignonieae. This paper utilizes the vast
reau (12 species), Potamoganos (l species), and amount of information chronicled by Gentry in his
Phryganocydia Mart. ex Bureau (4 species) (Gentry, studies of Bignoniaceae and attempts to compile
1977, 1997: Gentry & Tomb, 1979; Tomb & Gentry, these sources on Roentgenia and Potamoganos into
unpublished: Hauk, 1997). a single treatment. The database established bs
~
Potamoganos is a monotypic genus of primarily entry's investigations has allowed production of
E yple 8 | \ B
lowland lianas with lavender to magenta, campan- detailed maps of geographic distribution (Figs. 3.
ulate-funnelform corollas (Fig. 2). Among genera of — 4). Distribution data presented here provide an im-
tribe Bignonieae with “Cydista-type” flowers, it is portant foundation for future investigations of the
! This paper is number 11 of the о INVITATIONAL SERIES, in acknowledgment of contributions to the
study of the Bignoniaceae made by Alwyn Н. Gentry
? | thank Peter Raven and the Missouri Botanical "Cards n for the — to conduct this research. In particular,
| am grateful to the late William G. D'Arcy, whose advice and guidance were an invaluable contribution to this work.
Susan A. Moore provided the illustrations. Financial support was — by the National Science Foundation (grant
DE В- 950927 О).
! Department of Biology, Denison University, Granville. Ohio 43023. U.S.A.
ANN. MISSOURI Bor. GARD. 89: 77-87. 2002.
Annals of the
Missouri Botanical Garden
exine
Pollen
1997; Gentry
44.
Pollen aperture
Seed
Fruit
surface
Ovule
organization
Corolla
pubescence
Nectar
disk
arms
Phloem
Morphological/anatomical characters and character states for Roentgenia and Potamoganos and three genera putatively closely related (after Gentry, 19
Tendrils
& Tomb, 1979; Tomb & Gentry, unpublished).
Table 1.
taxonomy and systematics of Roentgenia and Po-
ate
ate
ale
ate
ate
tamoganos.
reticu
reticul:
HISTORY AND SYSTEMATICS
In 1916, Urban described Roentgenia and trans-
ferred Cydista bracteomana К. Schum. ex Sprague
into the genus as its sole species. Urban considered
Roentgenia distinct from Cydista based on the trifid
tendrils and plurisuleate pollen of the former
(Sprague & Sandwith, 1932). Sprague and Sandwith
transferred Arrabidaea sordida Bureau & K.
inaperturate/pericolpate
3(—4)-colpate
Inaperturate
inaperturate
Schum. to Roentgenia in 1932, after examining ad-
G
3-colpate
ditional collections more complete than the holo-
type. Macbride (1961) questioned the use of pollen
characters alone as generic markers, and noted that
м. other characters, for example simple vs. trifid ten-
bialate
bialate
corky
drils, that separate Roentgenia from Cydista are not
stable in Bignoniaceae. Gentry (1974a) acknowl-
edged that the primary features distinguishing the
two genera, tendril and pollen, might not be suffi-
Y
cient to retain Cydista and Roentgenia as distinct.
Gentry (1978: 262) wrote that Roentgenia is “barely
separable from Cydista,” although he treated them
non-echinate
non-echinate
non-echinate
echinate
as distinct in his floristic works (Gentry, 1977,
1978, 1983, 1993, 1997).
Gentry (1974a: 880) reported that the distinction
between the trifid tendril of Roentgenia and the
2-seriate
2(—4)-seriate
2(4)-seriate
}-seriate
simple tendrils of Cydista “breaks down” in a col-
lection of R. bracteomana (Seibert 2146; not ex-
amined by this author), However, the putative pres-
ence of anomalous tendrils in a single specimen
does not provide a strong argument for treating
Roentgenia and Cydista as congeneric. From a cla-
distic perspective, the perisyncolpate pollen. linear
linear-trisect. bracts and bracteoles, and trifid
glabrous/lepidote
lepidote
lepidote
lepidote
tendrils of R. bracteomana and R. sordida provide
synapomorphies that indicate they are sister spe-
cies. Because there is no evidence, either morpho-
logical or molecular, that provides a sound basis for
absent
absent
present
absent
absent
combining the two genera, this treatment. retains
Roentgenia as distinct. from Cydista. Whether
Roentgenia arose from within a paraphyletic Cydis-
di со pu е ta (or conversely), or as sister to Cydista or another
closely related genus, is not clear at this time. Only
future, detailed morphological and/or molecular in-
vestigations will be able to address rigorously the
monophyly of Roentgenia and Cydista, and the re-
bi-trifid
simple
simple
trifid
lationship of these two genera to the presumably
closely related Clytostoma, Phryganocydia, and Po-
—
. F. W. Meyer described Bignonia microcalyx
in 1818, and Sandwith (1937) transferred it to Po-
tamoganos, noting its affinities to Cydista and
lytostoma
Roentgenia
Cydista
Phryganocydia
Potamoganos
4
Roentgenia. However, Sandwith (1937) asserted
C
Volume 89, Number 1
002
auk 79
Roentgenia and Potamoganos
that the presence of a nectariferous disk distin-
guished the taxon from both Cydista and Roentgen-
The cupular calyx and “Cydista-type”
Potamoganos (Fig. 2A) led Gentry (1978) to con-
clude that Potamoganos is allied to Cydista and
Roentgenia. However, Potamoganos differs. from
these genera in having a well-developed nectari-
ferous disk (Fig. 2C). and lacking glandular fields
1978).
other genera, Clytostoma, Roentgenia, and Phry-
[presumably on the calyx] (Gentry, Three
“Cydistatype” corolla but
1978). However.
Phryganocydia has a long, spathaceous calyx, not
ganocydia, possess the *
ack a nectariferous disk (Gentry,
the cupular calyx characteristic of Roentgenia and
Cydista.
The 3-colpate pollen. of Potamoganos differs
from the inaperturate pollen of Clytostoma. Phry-
ganocydia, and most species of Cydista (Gentry &
1979).
with Cydista-type corollas that has aperturate pol-
Tomb, Roentgenia is the only other genus
len (Le., colpate), although perisyncolpate pollen is
known in four of the six species of Cydista (C. de-
a]
cora, С.
y
heterophylla, and C. ae-
1979).
Roentgenia pollen is verrucate to scabrate. and un-
diversifolia,
quinoctialis; Gentry & Tomb,
like the medium-reticulate surface found in Pota-
moganos pollen (Gentry & Tomb, 1979). Phrygan-
ocydia has finely scabrate pollen that differs from
the medium-reticulate type of Potamoganos, Cly-
tostoma, — most species of Cydista (Gentry &
Tomb,
The — numbers of Aoentgenia and Po-
Goldblatt and
Gentry (1979) reported chromosome numbers of 2n
40 in 21 of the 23 genera of tribe Bignonieae
including Cydista and Clytostoma.
tamoganos are not known. However.
they surveyed,
and it is likely that Roentgenia and Potamoganos
have a similar chromosome number.
DISTRIBUTION
Collections of R. bracteomana extend from south-
ern Colombia to northern Bolivia, with some exten-
sion into north-central and western Brazil (Fig. 3).
Collections of R.
Guyana,
sordida occur from eastern. Ven-
ezuela, Suriname, French Guiana. and
northern Brazil. with a single collection from Ama-
zonas. Brazil (Fig. 3)
Collections of Potamoganos microcalyx occur
from south-central Venezuela, Guyana, Suriname.
and the northernmost portions of Brazil (Fig. 4).
western Suriname, Р, microcalyx grows in lowland
forests (0-60 m) and is present in many moist for-
ests and riverine habitats (pers. obs.).
flower of
The surface of
PHENOLOGY
Roentgenia and Potamoganos are closely related
to a group of genera (Cydista, Clytostoma. and
Phryganocydia) that Gentry and Tomb (1979) sug-
gested constitute a natural group because they
share similar pollen morphology, a tendency toward
“multiple bang" flowering phenology, and the ab-
sence of a nectariferous disk. "Multiple bang" spe-
cies have numerous, synchronized, short flowering
days) that may occur at any time of
1974b). The absence of a nectar-
iferous disk (and presumably nectar), coupled with
periods (ca. 3
the year (Gentry,
conspicuous visual and olfactory attractants. indi-
cates that pollinator deceit may be the ultimate pol-
lination strategy (Gentry, 1974b). The short, re-
peated floral bursts may serve lo lure novice
pollinators that effect pollination through visits to
only a few flowers, after which they seek a more
ample nectar source (Gentry, 1974b). Visits by po-
tential pollinators are infrequent. presumably be-
cause pollinators learn quickly that the flowers offer
no nectar reward (Gentry, 1974).
Fertile collections of Roentgenia bracteomana
and R. sordida are known from throughout the year
and are consistent with a "multiple bang" pollina-
tion syndrome. However, little is known about the
pollinators of these two species, and more detailed
examinations of flower production are necessary to
document reproductive mode.
The reproductive biology of Potamoganos is not
documented, but the close relationship of Pota-
moganos to “multiple bang” genera suggests a pre-
disposition to a "multiple bang" strategy. However.
the nectariferous disk in Potamoganos microcalyx
(Fig. 2С) could signal a departure from the “mul-
tiple bang” strategy. The flowering specimens ex-
amined for this treatment document that PF micro-
calyx flowers in February, April, May, and October.
Thus. the phenology data available are consistent
with a “multiple bang” syndrome for P. microcalyx.
MATERIALS AND METHODS
Gentry compiled a private database of label in-
formation from herbarium specimens he collected
and from specimens at other herbaria that he ex-
amined personally. Gentry’s database has been in-
corporated into the Missouri Botanical Garden da-
TROPICOS,
also contains label information for all Roentgenia
tabase-management system, which
and Potamoganos specimens housed at MO. All
types were assumed to have been seen by Gentry
unless otherwise noted. Gentry did not always des-
ы 99 99 66+ ee
ignate types as holotype, isotype, or “syntype,
and the designations presented here are based upon
Annals of the
Missouri Botanical Garden
inferences drawn from Gentry’s work and the orig-
inal literature; these type designations were not
based on personal verification of specimens at the
various herbaria. An index to numbered exsiccatae
is provided in Appendix 1.
Data used for mapping and phenology were down-
loaded from TROPICOS, <http://mobot.mobot.org/
W3T/Search/vast.html>. For records with no latitude/
longitude coordinates in TROPICOS, approximate co-
ordinates were obtained from gazetteers produced by
the U.S. Board on Geographic Names, Office of Ge-
ography, Dept. of the Interior. Distribution — were
produced using the computer program VE MAP
1.51 (C. H. Culberson, Newark, DE, оо
TAXONOMIC TREATMENT
Ber. Deutsch. Bot. Ges. 34: 747
Roentgenia bracteomana (K.
Roentgenia Urb.,
916. TYPE:
Schum. ex Sprague) Urb.
Lianas; stems terete to subtetragonal with 8 phlo-
em arms in cross section; branchlets terete to sub-
tetragonal with interpetiolar glandular fields lack-
ing and transverse interpetiolar ridges present or
absent, the younger branchlets striate, glabrate to
lepidote; pseudostipules caducous to somewhat
persistent, foliaceous, chordate-orbicular, glandu-
lar, glabrate to lepidote. Leaves opposite, estipulate,
petiolate, bifoliolate with the terminal segment of-
ten modified into a trifid tendril; petioles and pet-
iolules terete to slightly flattened, glabrate to pu-
berulent or sparsely lepidote; ultimate segments
entire, chartaceous, marginally plane; venation bro-
chidodromous; midrib prominent with secondary
veins pinnate; axils of secondary veins lacking
glandular fields. Inflorescences elongate, axillary or
terminal racemes, branched or unbranched, sever-
al- to many-flowered; rachis and peduncles terete
to subtetragonal, puberulent to lepidote, and brac-
teate, the bracts linear to linear-trisect, each seg-
ment linear, persistent to caducous, glandular or
eglandular, puberulent to caducous; pedicels terete,
puberulent to lepidote; bracteoles linear to linear-
trisect, persistent to caducous, glandular or eglan-
dular, puberulent to lepidote. Flowers ovoid in bud:
~
calyx campanulate-funnelform, glandular or
eglandular, costate or ecostate, glabrate to lepidote
or puberulent, the calyx margin usually split, api-
cally truncate except for five minute teeth; corolla
zygomorphic, tubular-funnelform, white to magenta
with purple markings and the tube yellowish, gla-
brate to lepidote externally, pilose pubescence at
the base of the filaments; corolla lobes 5 (2 upper
and 3 lower), short-orbicular, the inner surface gla-
brate to lepidote, the outer surface lepidote: sta-
mens 5, fertile stamens didynamous with the fifth
stamen modified into a staminode, all adnate to the
corolla; fertile anthers with two spreading thecae,
included, glabrate; disk absent; ovary cylindrical,
lepidote, the ovules biseriate in each locule; stigma
included, bipartite, the divisions laterally flattened.
Fruit a septicidal capsule dehiscing parallel to the
septum, linear-oblong, the valves compressed and
not conspicuously thickened, the margins not se-
rially constricted, drying brown, the midline not ev-
ident, the surface rough, plane and glabrescent or
sparsely lepidote to puberulous; many-seeded;
seeds flattened, oblong, bialate, the body distinct
and bipartite.
KEY
1. Bracts 10-15 mm long; bracteoles 5-10 mm long,
extending beyond the base of the calyx
R. bracteomana
I’. Bracts 1-5 mm long; bracteoles 1-3 mm long,
not extending to the base of the calyx R. sordida
TO FLOWERING SPECIMENS
1. Roentgenia bracteomana (K. Schum. ex
Sprague) Urb., Ber. Deutsch. Bot. Ges. 34:
747. 1916. Cydista bracteomana K. Schum. ex
Sprague, Verh. Bot. Vereins Prov. Brandenburg
1908: 121. 1909. TYPE: Brazil. Amazonas:
Victoria, Rio Jura, May 1901, Ule 5497 (ho-
lotype. B not seen; isotypes, HB not seen,
4
not seen, MG not seen). Figure 1.
Lianas; branchlets drying light brown to gray, te-
rete to subtetragonal, with interpetiolar transverse
ridge evident but inconspicuous, the lenticels not
evident or inconspicuous; pseudostipules caducous,
8-11 X 15-18 mm. Leaves 10-37 ст long; petioles
0.5-6.0 cm, terete to slightly flattened, glabrate to
sparsely lepidote; petiolules 1.0—5.0, terete, glabra-
te to lepidote; ultimate segments (7)10.5-28 X 4.5
X 14 cm elliptic to ovate-elliptic, densely lepidote
when young, in older leaves glabrescent to sparsely
and inconspicuously lepidote, especially along the
basal portions of the veins beneath, apically acute,
basally acute to obtuse, with 5-6 principal vein
pairs. Inflorescences axillary racemes to 30 em long,
unbranched, several- to many-flowered; rachis and
peduncles terete to subtetragonal, lepidote, and
bracteate, the bracts trisect, each segment linear,
10—15 X
pedicels 5—
2—4 mm, persistent, glandular, lepidote;
7 mm long, lepidote; bracteoles 5-10 X
1-3 mm, persistent, glandular or eglandular, lepi-
dote, extending beyond the base of the calyx. Flow-
ers: calyx 5-10 X 4—7 mm, glandular or eglandular,
costate, the outer calyx surface sparsely lepidote,
the inner surface glabrate, the margin intact or split
1/4—1/2 the length of the tube, the valves apically
Volume 89, Number 1
2002
Hauk 81
Roentgenia and Potamoganos
Figur
~
after Plowman «€ Schunke V. 75(
acute to truncate except for 5 minute teeth: corolla
exserted 30-35 mm beyond the calyx lip (35—40
mm total length), 2-3 mm wide at the calyx lip;
corolla lobes 15 X 10 mm; fertile stamens 9 or 10
mm long. inserted ca. 10 mm beyond the corolla
tube base. the staminode 2 mm long, inserted 2 mm
proximal to the stamens; ovary 2-3 mm long: style
ca. ЗО mm long. Capsule 27-60 X 2.0-2.5 mm,
I. Roentgenia bracteomana.—A. Leaves with inflorescence, after R. Burnham 1563.—B. Fruit with seed.
)3.—€C. Seed, after Plowman & Schunke V. 7503.
glabrescent; seeds flattened, 2.0-2.3 X 5.5-0.2 cm.
1-2 mm thick, oblong, bialate, the body ovoid and
bipartite. Figure: Gentry (1993, fig. 71—06).
Distribution, elevation, and habitat. Known
from southern Colombia to northern Bolivia with a
few collections from western and north-central Bra-
zil (Fig. 3). Collections are reported from 150 to
Annals of the
Missouri Botanical Garden
870 m, in premontane wet and tropical wet forests,
on terra firme forests or in riverine habitats.
Phenology. | Flowering collections of Roentgenia
bracteomana are from: January (10), February (4),
March (4), April (1), May (3), July (1), August (3),
September (1), October (7), November (1), and De-
cember (1). Fruiting collections are from: February
(1), March (1), April (1), May (3), June (4), July (3),
September (2), and October (3).
Representative үөү ОМВ1А. Caqueta: 41
km N of Florencia, 1060 m, e 9169 (MO). Putu-
mayo: Selva higrofila del Río Putumayo en las márgenes
del afluente izquierda La Concepción, 225 m, Cuatrecasas
10834 (COL). ECUADOR. Napo: Armenia Vieja at Río
Napo, 12 km SW of Coca, Lugo 2703 (GB, MO). Morona-
Santiago: Taisha, Río Guaguaine, 500 m, Cazalet & Pen-
nington 7558 (K, IS). Pastaza: Río Bobonaza, Quil-
loallpa below Montalvo, 02°10'S, 76?53'W, 300 m.
Øllgaard et al. 34586 (AAU, MO). Santiago-Zam«
banks of Río — mi, 500 m, Cazalet & PDA
NY, US). Sucumbios: without exact locality,
240 m, 00°08'S, 76°22'W, Carlos E. Cerón & Judith Ayala
9469 фе ГУ A a 'hipe: without exact loc — ty,
930 m, 04°18'S W, W. Pe l. Varg
Freire 8717 (МО). И ч ›пав: ONF :
Woytkowski 5637 (MO). Huánuco: Codo de Pozuzo, al-
luvial fan floodplain of Río Pozuzo, 10°15'S, 74°55’ W, 300
m, Foster 938 0). Loreto: Alto Amazonas, Río Pas-
taza, Lago Rimachi, 04°20'S, 76°35'W, 200 m, Díaz S. &
Ruiz 924 (AAU, MO). Madre de Dios: 39 km SW of Pto.
Maldonado, Laguna Cocacocha, d 69*20' W, Smith
& Shuhler 400 (MO). Puno: „пове Бе
Насі LOS,
ween ps ү andamo
and Río Guacamayo, 13°30'S, 69°50’ Е 400—6 1, Gen-
try et al. 76941 (MO). San Non Fundo Pts Pa
Mariscal Caceres, Tocache Nuevo, 450 m, Even
11788 (F). Ucayali: vicinity of LSU base c amp. x —
Shesha, ca. 65 km NE of Puc PE 08°02'S, 73°55'W, 25
m, Gentry & Diaz 58505 (MO). BOLIVIA. Beni: Km |
carretera е ae a Colegio Técnico Agro-
pecuario Río Colorado, 14°50’'S, 67°05'W, 230 m, Smith
et al. 14115 (MO). La Paz: Abel Iturralde Province, Alto
Madidi across from mouth of Río Enlatagua, 13?35'S,
68°46'W, 280 m, Gentry & Estensoro 70344 (MO). BRA-
ZIL. Amazonas: 54.
Maranhao: Rio Xingu, 04°49'S, 52°31 W, Balée 1958
(
NY).
Flowering material of Roentgenia bracteomana is
distinctive because of the prominent bracts and
bracteoles; no species of Cydista or Clytostoma
have similar-sized bracts or bracteoles. The lack of
conspicuous bracts and bracteoles in R. sordida,
coupled with a largely allopatric distribution,
makes confusion of the two species unlikely. Veg-
etatively, trifid tendrils distinguish Roentgenia from
Cydista and Clytostoma, but sterile material lacking
tendrils may be easily mistaken for species of these
closely related genera.
2. Roentgenia sordida (Bureau & K. Se pes
Sprague & Sandw., Bull. Misc. Inform. Kew
1932: 91-92 1932. — sordida Bu-
reau & K. Schum., in Mar s. 8 pt. 2,
fasc. 118: 30-31. 1896. TY PE: Guyana. Upper
Rupunini River, Schomburgk 1296 (holotype,
B not seen, presumed destroyed).
Arrabidaea — Sprague, Bull. Herb. — 28
5): 373-374. TYPE: Suriname. Sipaliwini: —
Saramacca River, Pulle 170, 495 liane U not
seen).
Lianas; branchlets drying tan to gray, terete, gla-
brate to puberulent or lepidote, with transverse in-
terpetiolar ridge evident but inconspicuous, the
lenticels not evident; pseudostipules + persistent,
5-8 X 6-7 mm. Leaves 11-27 cm long;
petioles 1.2-5.0 cm, terete, puberulent; petiolules
0.6-3.0 cm, terete, puberulent; ultimate segments
8-22 X 3-15 cm, elliptic to ovate, glabrate to lep-
idote (especially on young material) or puberulent,
especially on basal portions of the main veins, api-
foliaceous,
cally acute to obtuse, basally obtuse to rounded or
inequilateral, with 5 to 6 principal vein pairs. /n-
florescences axillary or terminal racemes to 20 cm
long, unbranched, many-flowered; rachis and pe-
duncles terete to subtetragonal, puberulent to lep-
idote, bracteate, bracts 1—5 X 0.5-1 mm, cadu-
cous, eglandular, puberulent to lepidote; pedicels
3-6 mm long, puberulent to lepidote; bracteoles
linear (to linear-trisect), 1-3 X 0.5-1.0 mm, ca-
ducous, eglandular, puberulent to lepidote, not ex-
tending to the base of the calyx. Flowers: calyx 4—
7 X 3-6 mm, glandular, ecostate, the outer calyx
surface puberulent to lepidote, the inner surface
glabrate, the margin intact or split 1/8-1/2 the
length of the calyx, the valves apically obtuse with
5 minute teeth; corolla exserted (25)32—42 mm be-
yond the calyx lip ((29)36—49 mm total length),
1.5-3 mm wide at the calyx lip; corolla lobes 12
X 12 mm: fertile stamens 8 or 16 mm long, inserted
9 mm beyond the corolla tube base, the staminode
2 mm long, inserted 1-2 mm proximal to the in-
sertion of the fertile stamens; ovary 4 mm long;
style ca. 25 mm long. Capsule 27-34 X 2 mm,
sparsely lepidote to puberulous; seeds not ob-
served. Figures: Gentry (1983, fig. 39; 1997, fig.
Distribution, elevation, and habitat. Known
from eastern Venezuela, Guyana, Suriname, French
Guiana, and northern Brazil, with a few, scattered
collections from western Brazil (Fig. 3). Collections
are reported from 350 to 2000 m, in non-inundated
moist forests, swampy mature forests, or white sand
areas.
Volume 89, Number 1
2002
Hauk 83
Roentgenia and Potamoganos
Phenology. Flowering collections of Roentgenta
—
5
sordida. are from: January (5), March (1), April (1
May (3), July (1), September (1). October (6). No-
vember (7). and December (4).
specimen examined was collected in March.
The sole fruiting
Representative spec m BRAZIL. Amapa: Mpio. de
49'N. 51?23'W, S. Mori & К. Souza 17263
. Amazonas: São Pa ulo de Olivenga, basin creek of
Sm Krukoff 8623 (BM, F, MO. NY, U). Maranhão:
Monção. P.I. Guaja, Rio Turiaqú, 03°07'S, 4670' W, Balée
3517 (MO). Para: Campus of IPEAN, Belém, Gentry
13077 (MO). FRENCH GUIANA. Approuague, en amont
de Crique — Oldeman 2303 (САҮ). Cayenne: Mon-
tagne de Kaw, Piste Roura/Kaw, 04°33'N, 52°09" W, 350
m, PX 2926 (MO). GUYANA. East Berbice: margins
of Berbice pes S of New — 06?00'N, 57°43'W
Maas et al. 5573 (U). Mazaruni: Upper Mazaruni jd
Mapa 170 m. Tillett & Tillett 45685 (MO). S
. Nickerie: right bank of Corantijn River, N P
Kaboe rie € E. ‚ Heyde 449 (MO). Para : Jodensavanne- Ma-
J 'oene, 02?
Ne
ri-Weg and de Crane-
Weg. W of Le pee in one sec, an on white sand, Heyde
596 (MO). VENEZUELA. Bolivar: Rio Paragua, hasta 12
vueltas т la boca del Rio Топого, 06%03'N
O3°4AT'W, 17 — 10320 (MO). Delta Amacuro:
E * Rio сш El Palmar, Gentry & Berry 14938
О). Mérida: Marcano Berti, Luis 199 (MO, VEN).
When intact tendrils are not present on a spec-
imen, the short bracts and bracteoles of Roentgenia
sordida make it particularly difficult to distinguish
from the sympatric Cydista aequinoctialis (L.)
Miers, although the latter can usually be distin-
guished by dark glandular fields (when dried) in
the axils of the secondary veins beneath; these
elandular fields were not observed in Roentgenta.
Gentry (1978: 27
paniculate inflorescence and overall leaf appear-
19) suggested that the racemose-
ance of Cydista lilacina A. H. Gentry “links” the
genera Cydista, Roentgenia, and Clytostoma. How-
ever, C. lilacina is distinct from other Cydista spe-
cies in its curved bud apices and 4-seriate ovules
(Hauk, 1997), and these two features serve to dis-
lilacina from К.
persistent pseudostipules and echinate
y
tinguish C.
liad-like,”
fruit of Clytostoma are distinct from the foliaceous,
sordida. The “brome-
caducous pseudostipules and rough (but plane)
fruit of R.
Mey.) Sandw. is partially sympatric with R. sordida
sordida. Potamoganos microcalyx (С.
and has trifid tendrils, but P. microcalyx differs
from R. sordida in having a nectariferous disk.
—— Sandw., Recueil Trav. Bot. Néerl.
220-221. 1937. TYPE: Potamoganos mi-
Mene (G. Mey.) Sandw.
Lianas; stems woody with 4 phloem arms in cross
section, branchlets terete to subtetragonal with in-
terpetiolar glandular fields present and distal to the
node, the transverse interpetiolar ridges not evi-
dent, the younger branchlets striate and glabrate to
lepidote; pseudostipules caducous, subfoliaceous.
triangular, eglandular, lepidote. Leaves opposite, es-
tipulate, petiolate, 2- or 3-foliolate with the termi-
nal leaflet often modified into a trifid tendril; peti-
oles terete, glabrate; petiolules terete to sulcate,
elabrate to lepidote; leaflets entire, marginally
plane; axils of secondary veins lacking glandular
fields. Inflorescences elongate, axillary racemes, un-
branched, several-flowered; rachis and peduncles
flattened to nearly terete, glabrate to lepidote, brac-
teate, the bracts minute, caducous, lacking circular
glandular fields, glandular-lepidote; pedicels flat-
tened, glabrate to lepidote; bracteoles minute, ca-
glandular-lepidote. Flowers
ducous, eglandular,
ovoid bud; calyx cupular-campanulate, eglan-
dular, lepidote, the calyx margin intact, the apex
truncate except for five minute teeth: corolla zygo-
morphic, campanulate-funnelform, lavender to ma-
genta, glabrate to sparsely lepidote, with an inner
ring of pubescence proximal to the insertion of the
stamens; corolla lobes 5 (2 upper and 3 lower),
short-orbicular, the inner surface glabrate, the outer
surface glabrate to lepidote; stamens 5, fertile sta-
mens didynamous with the fifth stamen modified
into a staminode, all adnate to the corolla: fertile
anthers with two spreading thecae, included, gla-
brous: nectariferous disk present; ovary cylindrical,
densely lepidote, the ovules 4-seriate in each loc-
ule; stigma included, bipartite, the divisions later-
ally flattened. Fruit not seen.
1. Potamoganos microcalyx (G. Mey.) Sandw..
Néerl. 34: 220-221. 1937.
Bignonia microcalyx G. Mey., Prim. Fl. Esseq.
211. 1818. TYPE:
type, GOET not seen), not seen by Gentry. Fig-
Recueil Trav. Bot.
Guyana, Meyer s.n. (holo-
ure 2.
Bignonia microc Ж var. acuminata Miq., in Flora 25(2):
427. 1942. TYPE: Suriname, (holotype?. U not
seen).
— — — Bureau & K. Schum., in Mart..
Fl. Bras. 8 pt. 2, fasc. 121: 146. 1897. TYPE: Su-
riname. prn exact locality, Wullschlaegel 1034
(BR not seen).
Lianas; branchlets drying brown to red-brown,
striate, glabrate, lenticels not evident: pseudosti-
pules triangular, 1 X 1 mm, eglandular, lepidote.
Leaves 15-25 cm long, once-pinnate, bifoliolate with
a terminal leaflet often modified into a trifid tendril:
petioles 3.5-11.0 ст long. glabrate to puberulent or
lepidote; petiolules 0.5-3.5 em long. glabrate to lep-
84
Annals of the
Missouri Botanical Garden
gure 2. a microc alyx.—
P r Miller & Hauk 9418.—
Tendril. after Miller & Pn 205
idote; leaflets 6.5-13 X 4.0-8.5 cm, elliptic to sub-
orbicular, plane, chartaceous, glabrate with a few mi-
nute glands scattered over the surface, apically acute
to obtuse, basally broadly acute to rounded, with 4 to
5 principal vein pairs, marginally plane. /nflorescen-
ces to 20 em long; bracts linear, 1.0 .5 mm;
bracteoles linear, 0.5-1.0 X 0.5 mm; flowers ovoid
in bud; calyx 3-5 X 5-6 mm, glabrate to sparsely
lepidote; corolla exserted 35-55 mm beyond the
. Inflorescence and leaves, after Miller & Hauk 9418.—B. Young —
Jetail — calyx, ovary, and nectariferous disk, after Miller & Hauk 9418.—
level of the calyx lip (40-60 mm total length), 2.5—
3.0 mm wide at the calyx lip, the corolla tube lep-
idote inwardly and glabrate outwardly; corolla lobes
18 X 15 mm; fertile stamens 12 or 18 mm long,
inserted ca. 5 mm beyond the corolla tube base,
the staminode З mm long, inserted З mm beyond
the corolla tube base; disk 1 mm tall; ovary 2-3
mm long; style 15-20 mm long. Fruit not seen.
Figures: Gentry (1983, fig. 37; 1997, fig. 407).
Volume 89, Number 1
002
Hauk
Roentgenia and Potamoganos
eJ
10N
0 >
10S
20S
30S
90W 80W ow
40W
Figure 3.
g Geographic
South America.
Distribution, elevation, Known
from south-central Venezuela,
and portions of northern Brazil (Fig. 4). Collections
are reported from O to 300 m in mixed lowland
and habitat.
Guyana, Suriname,
forest, often on lateritic soils.
Four flowering collections were ex-
amined: February (1), April (1), May (1), and Oc-
No fruiting solite ‘lions were seen or have
Р) е
H henology.
—
tober (1).
been documented.
— = ns. BRAZIL. Amapa: without ex-
act locality, 02703'N, 50°48'W, Ducke 1973 (MG
zonas: Boa Vista Road, 64 km N
60?08' W, Lowe 4190 (MO). GUYANA. Vicinity of ys
1 of Mazaruni and Cuyuni Rivers, 00730' М,
Station, junction
: Rupununi: Feat: of TM
50 40'W, Graham 307 (СМ).
Kanuku Mts., Moco-Moco village, 03° ‚ 59°35 W,
100 m, Maas К Westra 3906 (MO). йай ы Malali,
Demerara River, 05°50'N, 58?15'W, de la Cruz 2726 (CM,
US); — Trail, 2: . Don Linden, 06?00'N,
5815'W, 20 m, Grewal 16 (U). SURINAME. Voltzberg:
without exact locality, Pulle 307 (U). Saramacea: Cop-
pename River, 3 km below base camp, 05730' М, 55°50 W.
Florschutz & Maas 2734 (О). Sipaliwini: Mamadam,
O04?40'N, 55?40'W, Florschutz & Florschutz 1155 (U);
Montibus р fluv. Kabalebo & Coppename Sin-
istera, Kabalebo River 1-5 km below airstrip, 03°50'N,
56°50'W, Florschutz & Maas 2340 (U); area of vidt bo
—
distribution of collections of Roentgenia bracteomana (circles) and R. sordida (triangles) in
O4°10'N, 57210'W, Heyde & Lindeman 192
— proje üt
MOX vicinity 2 Blanche Marie falls on the Nickerie Riv-
~ O4°45'N, 56°52'W, 60 m, Miller & Hauk 9418 (BBS,
vi o Fluv. Sarama ca,
— 04^20' N,
l ои
р 25'W,
Baud & Huber 15362 (MO): Río Hennes be S San
Fernando de — Cano Morocoto, 1 hour below San
Fernando, 04°10'N, 67°42'W, Gentry i al. 10947
Rio Cuao, Río ае ‘0, — 617740'W, 125 m,
guire & Politi 28436 (NY, US); Cerro Huachamacari. Río
Cunucunuma, 03748' N, da °46'W. 400 m. Maguire et al.
29973 (VEN).
The shape and color of the Cydista-type corollas
of P. microcalyx are easily confused with those of
the sympatric Cydista aequinoctialis (L.) Miers and
C. lilacina A. H. the
axes of P microcalyx are longer and more
Gentry. However, inflores-
cence
flattened than either Cydista species, and the calyx
is shorter and more cupular. The nectariferous disk
and trifid tendrils of P. microcalyx should readily
distinguish it from all Cydista Gentry
(1997) noted that the leaves of Potamoganos mi-
crocalyx are similar to those of Distictella Kuntze
but that the inflorescences
Mansoa A. DC.
species.
and Ceratophytum Pitt.,
and flowers more closely resemble
Annals of the
Missouri Botanical Garden
70W 60W 50W 40W
Figure 4. Geographic distribution of collections of Potamoganos in northern South America.
and Roentgenia. Like Potamoganos, Roentgenia ‚ In press. Bignoniaceae. In: z rra de Colombia.
but the nectariferous disk is
absent, and the bracts and bracteoles of P. micro-
has a trifid tendril,
calyx are not trisect, as is found in Roentgenia.
Literature Cited
Gentry, A. H. 1974a. Studies in Bignoniaceae 12: New
— species of South — an Bons eae.
— Bot. Gard. 61: 872—885.
)74b. Flowering phe — and diversity in
tropic a Bina eae. Е a 6(1): 6 ita
— ———. 1977. Fam. 178. Bignoniaceae. гъ Harling
B. — (editors), Flora of Ecuador. o ra Bot.
173.
T
. 1978. Bignoniaceae. /n: The Botany of the Guy-
ana Highland: А Mem. New York Bot.
Gard. 2
xu аний eae. In: Z. Luces de Febres & T
A. Ste — (editors), Flora de Venezuela. 8(4): 7
433. Fundación iie Botánico de Venezuela, Ca-
racas.
—————. 1993. A Field Guide to the Families and Genera
of Woody Plants of Northwest South America (Colombia,
Ecuador, Peru) with Supplementary Notes on Herba-
ceous Taxa. Conservation. International, „ишн
D.C. [Reprinted by Univ. Chicago — )o.]
: Bignoniaceae. Pp. 403—491 in: J. A.
ermark, А E. Berry & B. K. Holst tt €
Flora of the Venezuelan — Vol. 3. Missouri Bo-
tanical Garden Press, St. Lou
& A. S. Tomb. 1979 [1980]. Taxonomic implica-
= ons of Bignoniac eae palynology. Ann. Missouri Bot.
. 66: 750-777
Goldblatt P. & A. H. Gentry. i Cytology of Bignoni-
. Not. 132: 47:
1997, A review » th. > genus — (Big:
4 Ann. Missouri Bot. Gard 84:
J. F. 1961. A. eae. In: — i я ru.
is. e Hist., Bot. 3-103.
1818. кле э (a Fl. Esseq 211.
Sandwith, N. Y. 1932. Notes on — American Big-
noniace a я ueil Trav. Bot. Néerl. 34: 220-223.
——— 1937. Notes оп yo val jua 'an Bignoniaceae.
Reci isi] аң Bot. — 1: 220-223.
Spangler, R. E. € R. a 1999, е
analysis of Rims eae based on the cpDNA gene se-
quences rbcL and ndhF. Ann. Missouri Bot. Card. 86:
33-46.
Sprague, T. A. & N. Y. Sandwith. 1932. оог to
the flora of — al America: IX. Kew Bull. 1932: 18—
=
©
=
=
=
=
^
e
e
=
b
(chan, 1. 1916. Uber Ranken n: — de Bignoni-
aceen. Ber. Deutsch. Bot. Ges. 34:
APPENDIX ]
INDEX TO NUMBERED EXSICC,
Collections are listed КИ ally by the princ ipal
author, followed by collection number, and a boldface
number (1—3) that indicates the collected. All
specimens entered into TROPICOS were assumed to have
spec cies
Volume 89, Number 1
2002
Hauk
Roentgenia and Potamoganos
been examined by A. Н. Gentry. Specimens examined by
the author were limited to duplic ates housed at MO, and
are indicated by a 7". The coding of the species number
: Roentgenia aracteamana = — l, Roe —
- 2. еми microcalyx
Archer, há A. 8005 2: Aronson, J. 1039! 1 ; Ayala, E
et al. 3927
Balée, W. L. 1958 l, 1968 E 3517! 2: Beck. H. T. et
al мо! 2 — R. 897! 2; Black, G. A. 47-819 2:
aye. J. «€ Asanza. E. 30067!
ura, Д Ј.
T. zalet, Ww — 7558! 1; rá E. &
Ayala, J. — l; Cremers, G. "Croat, T. B.
ч
21259! 1: Cruz, J. S. de la 2726! 3; Petite J. 10834.
1
у,
~
Davidse, G. & O. Huber apne З; Díaz S.C. & J.
Ruiz 924! 1; Díaz S., C. et al. 1297! 1; dos Santos, G. el
Fi ‚1 ж 2 Ducke 1973 3.
!1.
Florschulz 1155
3: Foster, R. B
; Florse *
Р. & Maas 2340,
5 billet, c
; Florschutz,
„б 1: Foster, R. B. & В. “Ас д 12004! I: Froes.
R. L. 31509 2.
ntry, A. H. 13077! 2, 49133! 2: Gentry, A. H. & P.
Berry 14938! 2, 14970! 2: Gentry H. & Díaz. C.
58505! 1; Gentry, A. H. & L. Emmons 39392! l: Gentry,
A. H. € S. Estensoro 70344! 1: Ge . H. & R. Foster
70842! 1: Gentry, A. H. & N. MN 1. 57083!
l; Gentry, A. H. € P. Núñez 59559 1, 69435! 1, 69559!
1: Gentry, A. H. & J. Revilla p l; Gentry, A. H. &
K. Young 31820! 1: . H. et al. 9169! С 10917!
. 220604! 1, — ‚ 25849! 1, 26787! 1,
27287! 1, re 1, 46076! 1, 68670! 1, 76904! 1,
76941! 1: Graham, E. ч 3: Grewal 16 3: Grández. C.
& А. Chigi puma *
Heyde, M.
— n s
ashikat, V. 215€
раш, J. A k “Coello 1433! 1.
„Р, — 2: Killip, E. P. & A. C. Smith 28169
g, G. 2043! 2170! l; Kramer & Hekking 7450
2: Krukoff, B. A. A eds 8847 2; Kujikat 357 1.
Lindeman, J. C. 4050 2, 5008 2, 5197 2: Linder, D.
H. 80 2; Lowe, J. 410! 3; Londoño, C. et al. 124! 3; Lugo.
H. 2557! 1, 2570! 1, 2703! 1.
Maas, P. J. M. & L. Y. Th. Westra 3906! 3; Maas, P. J.
Gentry, A
)! "4 * 2:
— Isen,
ж N. M. & J. С.
„ et al. 21432 1: Hu-
et al. 5573 2: Maguire, B. & L. Politi 28436 3; Maguire,
B. et al. 29973 3; Marcano-Berti, L. 199! 2; McDaniel,
S. € B. Marcos 11000! I: Mexía, Y. 6320! 1: Miller, J.
S. & W. D. Hauk 9418! 3; Mori, S. et al. 17251! 2: Mori,
S. & R. ups 17263! 2.
. et al. 6283! 1; Núñez. P. 6149! S
Timaná 12149A!
Oldeman. R. A. 2303 2: Øllgaard, B. et a
‚ W. et al. 8717! 1; Phillips, O. et al.
Pires, J. M. & N. T. Silva 11042 2,
2: Plowman, T. & т chun ke 7503! 1: Prance. G. T. et i
Е ; Pulle, A. 170 2, 307 3, s 3.
11852! 1,
345% 6! 1.
329! 1;
‚М. 577 2; Schulz, J. 8371 2; Se Lus Е
: 5696! 1, 11788 E; Seibert, R. , >
MG8824 2; Smith, А. С. 2847! А с
‚ 14115! 1: Smith. S. F. & prin
et al. 620! 1, 1454! 1; Steege, Н. e
! 3; Stergios, B. 10320! 2; Stergios, B. & L. nir
e 12998! 2.
Tessmann, G. 3245 1; Tillett, 5. S. & Tillett 45085! 2:
1 M. & P. и 1 426! 1; Timaná, M. & O. Phillips
881! 1.
"s E. — l.
| wed й
Young. K. 25 I:
331!
p H. J. & D. A. Stratton 3! 1.
APPENDIN 2
INDEX TO SCIENTIFIC NAMES
{nemopaegma
cupulatum 83
Arrabidaea
yullet 82
sordida 78. 82
Bignonia
mi rocalyx 78, 83
microcalyx var. acuminata 83
Ceratophytum 85
Clytostoma 77, 78, 79, 82, 83
Cydista 77, 78, 79, 82, 85
aequinoctialis 79, 83, 85
bracteomana 78, 80
ecora 79
diversifolia 79
heterophylla 79
lilacina 83. 85
Distictella 85
Mansoa 85
Phryganocydia 77, 78, 79
Potamoganos 77, 78, 79, 83, 86
nicrocal yx 77, 79, 83, 84, 85, 86
Roentgenia 77, 78, 79, 80, 82, 83, 86
bracteomana 77, 78, 79, 80, 81, 82, 85
sordida 77, 78, 79, 80, 82, 83, 85
MOLECULAR DATA INDICATE Marcus Koch? and Ihsan A. Al-Shehbaz*
COMPLEX INTRA- AND
INTERCONTINENTAL
DIFFERENTIATION OF
AMERICAN DRABA
(BRASSICACEAE)!
ABSTRACT
The genus Draba — about 350 species — ү in the Northern Hemisphere, with some 80 species
described morphologic rally, the existing sectional
iong species of
in South and Central America. Although spec ies of Dr are well
classification is — controversial. American taxa e hib it enormous morphological differences even an
d the hypothesis that variation accum ul; ated during migration and i ien of American
"the ITS (internal transcribed spacer) regions of the
72 nien; 'an gue taxa and 6
Mi primarily to pru
n TI
the same section. We teste
Draba. The present “eme netic study i
nuclear ribosomal DNA and we chloroplast trn L-intron and spacer sequences froi
European Draba species. Resul some intrageneric groupings corre
and that only to a small degree do — findings agree with previous sec — classific ‘ation. Differences betw ‘S-
ed between some of the groups or sections
ITS sequence zu lastid DNA
y the American Draba
^
2
—
is based on analyses
ts suggest that s
and trnL-derived phylogenies suggest extensive genetic contact may have exis
and that this disjunction = tween К uropean and American — is de ш ы һу
sequences suggest | that “European” plastome types be more widely distributed among
species, perhaps et — transmissions of Eurasian n chloroplast types d American Draba. Additional sys-
tematic analysis demonstrates that the genus Erophila has to be integrated into Analysis on the tribal level
— the entire Draba complex to be close to European Arabis and Aubrieta. The data provide additional support for
evious assumptions that the existing tribal classifications of the Brassicaceae are mostly artificial and that the seg-
re sharin of Draba and Arabis into separate tribes or subtribes does not accurately reflect their phylogenetic relationship.
Draba.
Key words:
Brassicaceae, Draba, Erophila, molecular systematics, reticulate evolution.
The wide occurrence of polyploidy in vascular
plants reflects its importance to plant speciation.
The origin of polyploids and the mechanisms be-
hind the establishment of newly evolved popula-
tions and taxa are among the challenging questions
& Schemske, 1998;
1999; Sol-
1999), New combinations of favorable
in. plant sciences (Ramsey
Thompson € Lumaret, 1992; Petit et al.,
tis & Soltis,
genes may be more easily stabilized in polyploid
taxa, and the permanent coexistence of favorable
traits and characters from different parental lines
may be effectively preserved as fixed heterozygosity
(Soltis & Soltis, 1993). These speciation processes
are integral to the genesis and maintenance of plant
biodiversity. Recent studies of the overall genome
structure of hybrid and polyploid taxa provide new
insights about the dynamic nature of complete ge-
nomes either analyzed on the basis of artificial hy-
brids or by comparative mapping (Rieseberg et al..
Linder, 1999;
1994). In their comparative
1999; Rieseberg &
1998; Kowalski et al.,
genome analysis of some Brassicaceae, Acarkan et
al. (2000) showed that structural rearrangements
occurred with a significantly higher frequeney in
Lagercrantz,
polyploid Brassica L. than in diploid Arabidopsis
thaliana (L.) Heynh. or Capsella rubella Reuter.
This suggests that at the structural level a more
dynamic nature of the genome might be sustained
in polyploid taxa. than diploid ones, and this
might be a remarkable source for new genetic var-
iation upon which selective pressures can then act.
Draba 1..
and includes ca. 350 species, or about 10% of the
family total (Al-Shehbaz, 1987).
tributed primarily in the Northern Hemisphere, es-
is the largest genus of the Brassicaceae
The genus is dis-
pecially in the subarctic to arctic regions and al-
mountainous portions of the temperate
Nearly half of the
pine or
regions. taxa are found in the
! We are grateful to je Mitchell-Olds, Max-Planck-Insitute for Chemical Ecology, Jena,
the curators of GH,
. Harriman, and Victoria Hollowell for critically reading the manuscript and offering
the facilities to conduct th arch, and we thank
thanks go to Donovan Baile ry, Nei cil.
numerous suggestions for its improvemer
? Institute of Botany,
' Missouri Botanical Garden, Р.О. Box 299, St. Louis.
ANN. Missourt Bor,
Germany, for providing
О, and US for the loan of specimens. Special
University for hen ultural Sciences Vienna, Gregor-Me "s Str. 33, A-1180 Vienna, Austria.
Missouri 63166-0299, A.
GARD. 89: 88-109. 2002.
Volume 89, Number 1
2002
Koch & Al-Shehbaz 89
American Draba
New World and about 80 species are distributed in
Central and South America (Al-Shehbaz, 1987).
Schulz (1927, 1936) divided the genus into 17 sec-
tions (Table 1), but Fernald (1934) criticized this
artificial sectional classification largely on its mis-
leading and impractical keys. Further, Tolmachev
(1939) totally ignored Schulz's sections and recog-
nized 29 series for the 91 species occurring in the
former Soviet Union. No subsequent botanists have
attempted to subdivide Draba into infrageneric
taxa. For European Draba, most authors have fol-
lowed Walters (1964) who correctly used section
Draba instead of Schulz’s section Leucodraba DC...
because section Draba includes the generic type.
A few case studies have investigated the phylo-
genetic relationships among some Arctic Scandi-
navian (Brochmann et al., 1992a-d, 1993) and al-
pine European (Widmer & Baltisberger, 1999a, b)
species of Draba. Some general aspects of these
studies have been previously outlined (Brochmann,
1992) and can be summarized as: (1) allopolyploidy
is common, with a polyphyletic and polytopic origin
for some taxa; (2) gene flow across different ploidy
levels is possible and probably occurs in natural
populations; (3) complex evolutionary phylogenetic
networks have been demonstrated for section Draba
and some species of section Chrysodraba:; (4) re-
peated migration and colonization contribute to
—
л
—
complex distribution patterns; and not unex-
pectedly, items 1 through 4 are substantiated by
molecular studies with complex patterns of intra-
and interspecific variation observed (Brochmann et
al., 1992a—-d: Widmer € Baltisberger, 1990a,
1999b). Support also exists at morphological, cy-
tological, and ecological levels (Brochmann, 1993:
Brochmann et al., 19
Isozyme elec се and DNA analysis of the
nuclear and plastid genome have greatly increased
the possibility of detecting and distinguishing allo-
and autopolyploids, to trace paternal and maternal
genome lineages, and to document hybridization,
introgression, and reticulate evolution within poly-
ploid complexes. Several polyploid complexes in
the Brassicaceae have been characterized, includ-
ing those of the genera Microthlaspi К. К.
(Koch et al.. 1998b; Koch € Hurka, 1999). Draba
(Brochmann & — 1992; Brochmann, 1993:
Brochmann et al., 2a—d, 1993), Cochlearia L.
(Koch et al., 1998a, — Yinshania Ү. С.
Y. Z. Zhao (Koch & Al-Shehbaz, 2000), Cardamine
L. (Franzke et al., 1998; Urbanska et al., 1997),
Biscutella L. (König, 1998), Brassica L. and related
genera (Anderson & Warwick, 1999), and Nastur-
tium R. Br. (Bleeker et al., 2000).
In order to gain a better understanding of the
[=
phylogenetic relationships within a large polyploid
complex, we examined the sequence variation from
the internal transcribed spacer regions (ITS] and
ITS2) of nrDNA for American species of Draba
1995: Campbell et al., 1995), the
1991: Gielly & Ta-
erlet, 1994; van Ham et al., 1994; Koch & Al-
Shehbaz, 2000), and the cp trnL-trnF spacer (Ta-
berlet et al., 1991). The derived
phylogenies were then compared with traditional
(Baldwin et al.,
p trnL intron (Taberlet et al.,
molecular
concepts based on morphological data.
MATERIALS AND METHODS
PLANT PCR-
AMPLIFICATION,
MATERIAL, DNA EXTRACTION,
AND SEQUENCING
Leaf material for DNA extraction was obtained
from herbarium specimens (Table 1). 1 was iso-
lated from 25—50 mg of dried leaf material using
the NucleonPhytoPure Kit (Amersham Lifescience)
following the instructions of the supplier. DNA was
stored in 10 mM Tris-EDTA buffer pH 8.0 at
—20°С.
Double-stranded DNA of the complete ITS re-
gion, including the 5.8 S rDNA region, was ampli-
fied. by PCR using ITS
primers initially designed by White et al. (1990)
and modified by Mummenhoff et al. (1997a) for
ITS4. The PCR profile used to amplify the ITS re-
gion followed the following profile: a hot start with
min. at 94°C, and 35 cycles of amplification (1
min. 94°C, + 38°C, 45 sec. 72°C), final elon-
gation step for 10 min. at 2 C, and — at 4°C.
The 18F primer (5'-GGAAGGAGAAGTCGTAA-
CAAGG-3') is located at m 3'-end of the 18 S
rDNA gene and primer 25R (5'-TCCTCCGCTTAT-
TGATATGC-3’) is located at the 5'-end of the 25
S rDNA. PCR products were purified. using the
Boehringer PCR product purification. kit (Roche
Molecular Biochemicals). PCR products spanned
the entire ITSI, 5.8 S rDNA, and ITS2 region and
vere cycle-sequenced directly without cloning us-
35 cycles of symmetric
л
5 sec.
<=
ing the Taq DyeDeoxy Terminator Сус ‘le Sequenc-
ing Kit (ABI Applied Biosystems, Inc.) and the 18F
and 25R primers. Products of the cycle sequencing
reactions were run on an ABI 377XL automated
sequencer.
The trnL. (UAA) intron was amplified and se-
quenced by using the universal primers “e”
B49317, 5'-CGAAATCGGTAGACGCTACG-3')
located at the 3'-end of the trnL(U A A)5'-exon and
"d" (A49855, 5'-GGGGATAGAGGGACTTGAAC-
3’) located at the 5'-end of the trnl(UAA)3'-exon
(Taberlet et al., 1991). The PCR profile used to
amplify the trnL intron followed Koch and Al-Sheh-
—
Missouri Botanical Garden
Annals of the
90
€lOLPIAV — 8069F 149 IOTOLLAV 041роит] єє (OIN) 9222 Diueapipy зу ouvsig *eunuazay J "чоон vsoynoiunf (q
9969141۷ — €€69tLIV 80S9PIAV ә}ХЎио8әршоюцул) 18 pal GOLF saut) *e[onzouaA Чәче Y uopurT pupryounf q
PIOLPIAV — 21069PLIV — OOTtOTIIJV — 2]£2uoonum) c£ (SA) GOLF saui) *e[onzouaA "Чәче Xy Uepury puniyounf q
SIOLLIAV — 9069tLIV — OSPOPLAV Dqp4po]p-) 1 (OW) £90] sonang- zado] Y UDLA[ZMY *ve[onzouoA "Ydue]g Y uapurT saprojasapf q
8969t[4V €S69PIAV 9OCOLIAV DQDIPO]A cg (SA) €6€8 punjdsy *10peno'q ppom Dsuojxo “(q
8€£40c LIV ZZLOZIAV Dqpapoono] VU q6661 "9219qsn[eg Y 1ouipra, *pue[iozits ләп DIQNP "(T
OTOLLIAY — SO69FTAV 8SHObTAV Dqp4poj«] 0£ (OW) 69c] Paoyoarsoy 29 люиә]ц “sopensy J "чоон vssaıdap ‘q
8869t14V — £€69PLIV — S9tOor [AV DQUDApOs A7) c9 (OW) 82382 ито Y што "vs SEI) "Y 5 шор хә ‘HNN DIJOfiSUIP ‘q
OLSOPLAV DqD1posA147) ZOI (1880) ZE8Z руту “Y SN nny DYofisuap ‘q
69869t 14V — c£69tIJV — TOTOTLAV Dunjsouio[ 19 (OW) 2729 suijoy 9 ио ұт ABI "Y Y шор хә ‘HNN юуоЙәипә ‘q
6cO0LTLIV — 7689 149 1894V I (SA) 8992 Ј9ә]7) *eiquio[o7) euejues Y [o2ue| DUDSDIALIDNI "(T
LCV9T LAV Dqpapo]T 8с (OW) 52901 1m 12 упш; *n1ag J YOoH numpid&ao q
9cO0LVLIV = S689pPLIV — PPPOPIAV © (SA) II9FG О4әшоуу эр spsp2o4]0n7) *erquio[o7) əng ppnjdokao ‘q
0669VI4V — I£69PLIV — €8hOPTAV DQDApOSs A417) 09 (OW) 8£££ unusapus “VSN uos&eq DIMNA (T
I669p 149. O£69PLIV — Z8bObIAV D]]aquaq(q 6€ (OW) 1881 7? 12 шә “VS шецеле) Dijofisspao (T
IIO!tLIV — OL69FTAV — £OtOTF [AV Dqpapo]p:) ce (OW) 962€ әт “opens zin prjofindofuoo "qq
PZOLLIAY — L699PLIV — LbrOTIJV — 2]uoZopum) 8 (OW) EZZ} Lag `т|әпләцәд PIE `4 `$ Pprydouorys ‘q
LELOCIAV DQDIPOINY] SOL 46661 uodzioqsnpeg » J9UIptA "PUB LIZ IMG addoy DIDIYJULIDI `1
PLOOVLAV — LPOOPTAV ОО©ОР1АУ 6L (М) £0£ `P 12 qqnaz) “eIquiojor) zequaus-|V DUDIDAO (ү
OPPOVLAV a]&z uod onum) L (OIN) 69POÍ[ sonang 4 - zado] y up4o[ -2MY “B[ONZIUIA OTP el © 117127724٩ `1
€ZOLVIAV — 8689PIAV 6bt9blAV II (HO) 86£2 «оттү y Among шо? ZEQYPYS-[V PUDAD]IIDg q
COO9FIAV = 9S69FTAV — OOSOPTAV DGDAPONAY A 101 (QASO) 681€ PHNH "ven pjoonn “(J
69690t[4V — ZS69FIAV — SOSOPTAV DUNISOULO | r8 (SA) 6€6 4vao[ ‘nad "PPM s2proqpanp ‘q
LIOLTIJ4V — TO6OTPIJV — OS FOFTAV £c (OW) 291 DaouoojsoN 9 1Du2]4ç лорепә МЯН sapiosstqp ‘q
€£969t[IV — 8€69tIJV — ZISOPTAV sisdoziy FOI (661€ 188S0) ‘U's 22ffnos; *eujsny "| Seplozip ‘q
TELOZIAV sisdoziy 601 96661 "129219qsnpeg Y 19urpra, “рившәдим "т Seploz ‘q
DELOZIAV sisdoziy 801 46661 "9519qsn[eg Y 1oupra, “риерәдтм "| sapiozip (T
LZLOZIAV ICLOCLAV sisdoziy 101 96661 u2x19qsn[eg Y зәшргдү “риерәлим "| ѕәрто210 `q
T969tl1V — 2€69tIJV — IICOTLAV sisdoziy £01 (818£& ASO) "ws 4affnayy *Хиршләс) “| sepiozi ратл
I969t [41V — 0969tL4V PICOLIAV LL (OW) 6262 Sump $ uno "yes sui[[oy (SUIT[OY) »20750214ppnb °”)
c969tl41V = OS69FTAV — SICOFTAV L (OW) 8662 suvi]ng $ uno “Y SN SUI[OY (4&2) ү) nspjznop pjJarsoisn?)
19a0eds que поли! “que ¿SLI С 9199], dos ou “DY (шАполәр UN Lreq.2u) 10]99[[0» ‘nuno ПОХЕ |,
pue “YN Qu (2Z61) 41992
Ses ‘ISLI 0} BUIPIOIOB
UOTBOYISSE[O
[BPUOTVIIE
“HSH 3[qe1 әш uo p231eorpur SI иоцеив[ахә [euonippy "Apnis 19pun [еиәјеш əy} jo elep UOISSIIIY "| AFL
Koch & Al-Shehbaz
American Draba
Volume 89, Number 1
2002
€169T LIV OT69T [LAN I0€9T LAV $81] tsoro 08 (3) [S69] UPN Binden рушла 'dsqns жәл], njpu1ajnd 1
€OTOT [AV DQDA4posA4u7) OL (ON) ZIES UNIDA y YIO: MH "VG UOSAR puiiadsoisajd q
OCPOPLAY e (OW) ££49 *oiqifonf ap «png “Жыз zwquaus-[y sapioujupamauoopnosd (|
1809P LAV OFO69T LAV COT9T LAV punisouioj 69 (ON) 12126 ?pr] $ 19497) Y SN ABI) “Y Y шор рока “q
€002T LAV 9169T LAV 69V9T LAV 09701рордоцу Vy (OW) 6% ££ uouo]og ‘nag ABI "y nzuodoqoid “(|
900LP LAV CLO9P LAV 89v9V LAV DQUJpO]V) ©? (OW) ОРРР 2?puqopngy niod ZINYS 74 70 (COQ) оиро (|
LCOLTIAV TOGO LAV €tV9T LAV a|tAsuoziapuim) $ (OW) PEGE 1m 19 012104 *etquio[o?) доц 7 70) Muezpy-jjouued “(1
0869F LAV 169 LAV Tor9t LAV DQDADOS KAi[;) 12 (OW) 90669 1” 19 иом — apuqoe|y nuosivd (T
c969t LAV 6€69T LAV LOPOPLAV DQDAPOSAIY’) 99 (OW) ££Z62 SUjOY Y suiJoy "Vs "NooH puuadsozijo ‘q
c669T LAV 6c69T LAV L89 LIV. Dqu.pouapy 8S (OW) 22 70 19 007 45-9 ISO розни "(Jy
€869P LAV 8€69T LAV DOVODIAV DQDIPO iyd L9 (М) 161 ЧОН “led z|nuos 74 70) 21107070 (T
[L6O9TLAV 0€69T [AV EOSOPIAV Se ct Мей a ca (М) 161 ЧОН “led ZING ^q 'O %9ртоЈотупош q
( san]
L00LT LIV p169 I4V -og ) nqpapoono] It (OW) 2127 Diuoanap) XY oupsig "eunuasay "шв" DAUD) IDU '(]
8002 LAV €I69TP IAV LOVOTLAV DQDIPOPQDYY Ot (OW) F290] 70 12 ш Mig ‘J оон nuda ona ‘q
QOPOPLAV 6€ (OIN) £€9 P 19 ddna) "exquioor) aqu oun] t]
6002 T LJAV. cI69T 14V COPOVIAV siptsoio( 8€ (OW) 6ZOF Sog `®|әпләцәх Your] (00H) nupun (T
SLOLPIAV €069F LAV SCTOT [AV [с (OIN) 6/ZE] Vowopog *eriT[og zequoys-|y pupizpdp] '(
PpZLOZIAV DPVAGOINS ] 911 6661 OBIS у PUp "pue[rozis 19-79 pulpy] Y
OV Z0c LAV €cZ0cL AV DQDLPOINA] SLI 1666 | PASNE Y WUP "ривроәдим ‘Td- 4g Pulpo] Y
CoOo9T LAV 8c69T LAV O8POP LAV Dquiapouopy LS (OIN) 0c29c ?10]paoo4g `оәтхәр MAH sisuojnaof “q
T669T [AV LcO9T LAV Dquipouopy 9c (OIN) 9c [c 4210251 7021xX9]A MAH sisuoppaof ‘q
T9869F LAV LCo9Or LAV o8T9T LAV 99 (OIN) 6F62 197019 “Y 71 uosugof Y zun[y 1493901 q
€869T LAV 9€69T LAV 88F9T LAV DQDAPOING] c9 (OWN) €&2€E8 о 12 su]oy VS uossegq pj422u1 '(]
9869F LAV €€69T LAV L8P9P LAV Т9 (OIN) £9c€8 717 19 sui]oy VS supy 122020274 "(T
C669T LAV 9c69t LAV OLPOVLAV сс (OIN) А0692 MSNOpezy *021X9]N чоләрүегу szsuaz]ppni “(|
OLO9TLAV LS69T LAV TO0SOT LAV. DqU4pO])) €8 (OW) Fc Чїтш$-толләрү $ ADSUDY “opens Z[NYOS “A 70) DuDAajsuaY (T
9669T LAV CCOOTIAV 8L2v9v IAV Dqp4po]]&ugd ye (OW) Слер Орус *оогхәр 9U991 DUDA PY `Q
8cOLTLAV £689V LAV СҮҮҮОТЯҮ с (S) 0206 /22]) “erquojor) PID Y oen) пиәшшту (f
OLOZT LAV TL I69T LAV POPOVIAV Dqu1popp?) 9c (OW) 0££9 sojquing| 10penoj] 3 004 "gy “Y
L869bL1WV T€69T IdV 98v9v LAV 0901роэтәт £9 (OW) 269 vopzpog “WSN Чета 0/9409 ‘A
CIOLVIAV 6069T LAV COPOVIAV DQDAPO]D/) ve (OW) 878801 40140] Y 40]4n] “ao шү Y ооң «әт ү
дәве ча uout que ES LI c 31481 IIS оп "09V (WAUOLOR UID LIEU) 40]29[[09 "Anuno?) UOXBL
pue "YN QA (L261) Z[NYos
ос ‘ISLI 0] BUIPIOIOR
UOTJBOYISSe]O
[ечоцоэ<
"panurmuo?) `1 3IqeL
Annals of the
92
Missouri Botanical Garden
[€622€4V OLI 86/1122 Yosyaug *eujsny ләцо (T) Pua "3
ZS6LLEAV Ill 86/SS9% yosysug *euysny due] 4 "y vibpmyiods оруолу
L6690Yl4V — tc6OtLld4V — LLTOTLAV cc (OW) 6FIF 9394 zeyoupg ‘nəq zequeaus-[y :oopana (T
1960t14V — TC69tl4V — Z20C9TPLAV Dqp4popp:) 99 (OZOPECHSA) "ws uossumf '10peno'] MAH 22207014 `q
8669tLIV EZ69Þ1AV 9LPOLIAV [€ (OW) #261 uosuozuof “eunuad1y z[nuos ^q 70) SSUIUDUNIN] q
OZLOZIAV DQqnD4poorio'] ELI 46661 193195099 » J9UIptAA "pue[rozit ^o aey) DSOJUIWO] `1
[PLOZLAV CZLOZIAV Dqpapoone] ZU 46661 "9219qsnpeg P 1oupra ‘риерәдим Алгу DSOJUIUO] ‘q
9:69t14V — SPOOPIAV — SOFOFTAV GL (OW) ZIZP ssoy Y pjoyosoy “y Sn VU Y suroy »7DjJaqumqns q
ZLO9VPIAY — 61694 149 COCOTLAV [9 (М) OS] uosauf ‘1openoy ‘ZOIN], DSOJÁIS "(T
97021 AV Dqpapoonog 901 96661 19319450 [89 » тэшр yooy SLIDJAIS `1
82.69tL1V — €P69FTAV 9619 14V Dqpapo]]&uq Ji (OIN) 99 4494 “W'S хеле) "V Ddap2oido4js (T
LLOOVIAV — PROOPIAV — Z26T19V LAV bl (OW) 8911 120212ng| X uvuipoos) үсү uos&eq Y opuqoe|y 2£2]pupis (T
SPPOrlAV pqvipouspy 6 (HO) 997€ aanidg “1openoy ppam pupaonads (T
6669t[4V ZZ69WIAY — €ZTOTLAV D(DIPO]D/) 0€ (OIN) 08001 2975/99 Y p103]10 чорепәҷ Зүг) ѕиәриә ‘q
0002tL4V — IZ69PIAY — FZ2TOPLAV pqvipopqnuy 6b (OW) 9ETET vowojog ‘eog ррәдү 814270105 "(T
SZOLPIAV — Ө6Ө8Ө1ЇЯУ — SPPOPTAV D(DIPO]D/) 9 (SA) SEZZ 4onvqaoqag ‘nəd z[nuos ^q “O P1111108 ‘q
LOOLPIAW — Oc69tLIV — £L TOP LAV Sb (OW) 192 /22]) ‘eıquo[0) josury x PuPjues Damas '(
CO0LtI4V — 6169tLJV — cLTOFLAV pqvipopqouy 2% (OIN) 2206 uouojos ‘ern!{og `ррәдү umaopndoos (T
OZOLFIAV — LO69PIAV ES POPTAV Dqp4popquiy cl (HD) ££09 4әпюфләфәдү “Ned z[nuos “4 O Maisnyos (T
[ZOLPIAV — 0069 149 cCTOFLAV bl (H9) SOLE ¡22 uUDY “erquiojor) [93uey у PURJURS 401101 ‘q
ZZOLPIAY 6689bTAWY — ICTOTLAV el (HO) V-£c£z &oiifonf 2 Хороод *eiquio[o-) zequaus-[y pupanoppa ү
6LO9FIAV — ZPOOPIAV — S6POPTAV pursouio[ 4j (OW) 86r£ Yooquazyanyy “WSN р[ечлә] Cure) sunidas q
€269t[1V — 9rt69tL1V — O6FOFTAV Dqpapo]]^uqd 97 (OW) 8696F 4Y “VSN xnPAso(] DUISSISOUIDA (ү
ЄОО/Р1ЛУ — 8169FIAV IZPOPIAV D(DIPO]D/) 9t (OW) 0168 1” 12 илә “1openoy ‘zounj, DA ydouodd `q
POOLPIAV — ZI69PIAY OLPOPTAV ТЕТО] cp (OW) 9202 zmsumy “AO аца 4 орта q
6I0LÞIĄV ZO69PIAV — FCTOTLAV si&isojo( gI (OW) 0ЕОР әд `т|әпзәцәд —— zequaus-y aq 'dsqns "лолу руоштарта ‘q
Jaoeds "quad поди ил] ZSLI с 9149], 92s “OU "23V (шАполәр WINLIEGIoy) 101931105 ^ципог) UOXP],
рче ‘ұма (2261) 211925
єс "ISLI 0] BUIPIOIIR
UOTJBIYISSP] 9)
[EUon
'penunuo?) '[ [qe]
Volume 89, Number 1
2002
Koch & Al-Shehbaz
American Draba
93
baz (2000): a hot start with 5 min. at 94°C, and 35
cycles of amplification (1 min. 94°C, 45 sec. 50°C,
45 sec. 72°C), a final ср ов step for 10 пип.
at 72°C, апа storage at ¿R products were
purified and eyela-caquended as described for ITS
analysis using the amplification primer c and d.
The trnL(UAA)-trnF(GAA) intergenic spacer was
amplified using the following primers: the universal
primer “e” designed by Taberlet et al. (1991) (5'-
GGTTCAAGTCCCTCTATCCC-3') and a newly de-
signed primer trnF-IGS-rev (5'-AGGATTTTCAGT-
CCTCTGCTC-3'). Amplification, purification, and
sequencing were performed as described for the
trnl, intron.
DATA ANALYSIS
Outgroup selection. Тһе genus Draba has been
variously placed into the tribes Alysseae or Dra-
beae (Hayek, 1911; Schulz, 1936; Janchen, 1942;
Al-Shehbaz, 1987). However, as pointed out by Al-
Shehbaz (1987), the Alysseae are a poorly defined
tribe with about 40 genera (15 monotypic) and
some 650 species distributed primarily in the Ira-
no-Turanian regions. Recent
molecular studies (Price et al., 1994; Koch et al.,
1999a, 2000, 1999b) clearly indicate that nearly
all tribal subdivisions of the Brassicaceae are high-
and Mediterranean
ly artificial. Therefore, we used the aligned ITS data
set from Koch et al. (1999a) in which tribal struc-
tures were analyzed, and taxa from the tribes Lep-
idieae, Sisymbrieae, and Arabideae were consid-
ered to infer relative phylogenetic position of taxa
under investigation. We added sequence data from
Erophila verna (L.) Chev., D. araboides Wedd., D.
aizoides l., and Cusickiella — (Rollins)
110,
spectively, Table 1) to the ae described by
Koch et al. (1999a). We included C.
because it was originally described in — and
only recently it and D. douglassú А.
transferred to the newly established —
Rollins (Rollins, 1988). Data analysis, distance and
parsimony methods, were performed as described
in Koch et al. (19993) to get comparable results to
this. analysis.
Rollins (accession nos. ‚ 103, and 77, re-
quadricostata
were
The results were used to choose an
appropriate outgroup. Our analysis herein demon-
strated that the European Arabis L. and Aubrieta
deltoidea (L.) DC.
iella is. Therefore, European Arabis (A.
are closer to Draba than Cusick-
alpina, A.
bryoides, A. pumila, A. blepharophylla) and Aubrieta
deltoidea were selected as outgroups for the further
ITS study of Draba. Cusickiella was excluded from
all further analysis.
ITS
data. DNA sequences were aligned. by
hand. In addition to sequences from American Dra-
за. we included 6 ITS sequences from European
Draba from Widmer and Baltisberger (1999a. Table
1) and 6 sequences from Arabis/Aubrieta serving as
outgroups (Koch et al., 1999a), resulting in a total
matrix of 80 ITS sequences. Distance analyses were
performed with Kimura-2-parameter distances us-
ing PHYLIP software package version 3.57c (Fel-
1995). The neighbor-joining algorithm
was used for tree construction. Bootstrap analysis
(Felsenstein, 1985) was performed using 1000 rep-
licates. For all analyses gaps were coded as missing
senstein,
characters. The parsimony analysis began with a
subset of taxa comprising representatives from all
putative clades included in the overall analysis.
This analysis was performed to demonstrate relative
branch length and to estimate additional confidence
parameters. The second data matrix used was iden-
lical alignment to the first data matrix. Parsi-
mony analysis was performed with unordered Fitch
parsimony using PAUP* 4.0b2 (Swofford, 1999).
The branch-and-bound algorithm was used to find
maximally parsimonious trees. Bootstrap analysis
(Felsenstein, 1985) was performed using 1000 rep-
licates with the bootstrap search algorithm. Decay
analysis (Bremer, 1988) was performed in addition
to the bootstrap approach, in order to assess the
confidence that could be placed in the monophyly
of elades. Decay indices (DI) were estimated ac-
cording to Baum et al. (1994). For all parsimony
analyses gaps were coded as missing characters.
DNA sequences
were aligned by hand using the boundaries of cod-
trnL intron and spacer data.
ing regions as anchor points with which to begin
the We included 8 trnL
spacer sequences from European Draba (Widmer
& Baltisberger, 19991, Table 1) and 67 American
Draba for a total of 75 trnL sequences. Draba cu-
netfolia (acc. no. 61). D. platycarpa (асс. no. 69),
D. reptans (ace
84).
ingroup taxa (results from the ITS analysis), were
alignments. intron and
. no. 72), and D. araboides (acc. no.
which are only distantly related to remaining
used as outgroups. The ancestral position of these
four taxa with respect to the remaining Draba an-
alyzed herein was also confirmed by an analysis
using tral, intron and trnl-trnF spacer regions from
and Aubrieta deltoidea L. (GenBank
AY034180 and AY034181,
Distance analyses were performed for the
Arabis alpina L.
accession numbers re-
spectively).
total data set as deseribed for the ITS data. In order
to obtain a relative support for branching patterns
of representatives from all putative clades, a subset
of Draba taxa was investigated with Fitch parsi-
mony as described for ITS analysis. For this anal-
ysis
gaps were coded as additional binary (0/1)
Annals of the
Missouri Botanical Garden
characters. All gaps differing in length and —
were coded as separate binary characters. The dis-
tribution of gaps and its coding can be — al
http://homepage.boku.ac.at/koch/.
Amplification failed to yield ITS
1) and D.
jorullensis (acc. no. 56), and trnL intron and spacer
7), D.
Missing data.
data for Draba magellanica (acc. no. 4
data was missing for D. bellardii (acc. no.
spruceana (acc. no. 9), D. pseudocheiranthoides
(acc. no. 12), D. cryptantha (acc. no. 28), D. litamo
(acc. no. 39), D. pterosperma (acc. no. 70), and D.
densifolia (acc. no. 102). For Draba carinthiaca
105) and D. stylaris (acc. no. 106) only
ті, intron and spacer data were available, and for
(acc. no.
the two Erophila species no plastidic data were
available. Data matrices are available upon request
or can be viewed at http://homepage.boku.ac.at/
koch/.
RESULTS
OUTGROUP SELECTION AND TRIBAL RELATIONSHIPS
Our reexamination and analysis combining the
ITS data matrix (a sampling of 36 ITS sequences
from species across Arabideae, but excluding Dra-
ba) from Koch et al. (1999a) with sequences added
from Erophila verna, Draba araboides, D. aizoides,
and Cusickiella quadricostata (in capital letters,
Fig. 1) resulted in trees identical in topology to the
phylogenetic hypothesis presented in that paper.
The data matrix of 603 bp contained 342 invariable
characters. From the remaining 261 variable nu-
cleotide positions 163 sites were potentially infor-
resulted in 12
most parsimonious trees with a length of 612 steps,
mative. Fitch parsimony analysis
a consistency index (CI) of 48.9% TO s
excluded), and a retention index (RI) of 71.8%. The
strict consensus tree (not shown) out of Hid 12
most parsimonious trees is identical (except for Er-
ophila verna, Draba aizoides and D. araboides, and
Cusickiella quadricostata that are added herein) to
the previously published phylogenetic network
(Koch et al., 1999a).
to that presented by Koch et al. (19993) (Fig. 1).
Draba is closely related to the European Arabis L.
The distance tree is identical
and Aubrieta Adans., and it is positioned within an
(Fig. 1).
to the western United
"Arabideae core group" The genus Cus-
ickiella,
States,
molobus Tausch and North American Arabis. As
shown by Koch et al. (1999a), the North American
blepharophylla Hook. &
Arn. group and a few other species, are better treat-
which is endemic
was resolved with North American Hali-
Arabis, excluding the A.
ed in the genus Boechera A. Live & D. Live (Live
& Live, 1975). For subsequent ITS analysis of all
Draba taxa, European Arabis and Aubrieta were
used as outgroups, and Cusickiella has been ex-
cluded from all subsequent analysis.
ITS DATA
The analysis of phylogenetic relationships among
ingroup taxa comprises 80 sequences from
Draba and Erophila of which 74 sequences are
novel. Six ITS sequences from Arabis and Aubrieta
were used to root phylogenetic trees as outlined
above. The total length of the aligned data matrix
is 626 bp. From 283 bp within the ITS] region,
109 sites were variable (including 37 autapomorph-
ic changes); from 166 bp within the 5.8 S rDNA,
11 sites were variable (all autapomorphic changes):
and from 177 bp within the ITS2 region, 20 sites
were variable (all autapomorphic changes). This re-
sults in a total of 68 potentially informative sites.
Uncorrected pairwise sequence distances ranged
up to 10% within outgroup taxa (59 nucleotide po-
sitions), up to 9.39€ within ingroup taxa (57 nucle-
otide positions), and up to 12.3% among the whole
data set.
(acc. no. 33) was excluded from these
because of an unusual ITS tvpe that differed mark-
edly from the remaining Draba taxa (13.1 to 16.6%
Argentinian Draba funiculosa Hook. f.
calculations
sequence divergence or 82 to 104 nucleotide po-
sitions). However, in the phylogenetic analysis D.
funiculosa clustered among European Draba (Fig.
2). The whole alignment required 44 gap positions
of 1 nucleotide in length, and 2 gap positions re-
quired 4 (Arabis alpina, A. blepharophylla) and 14
(A. alpina L., Aubrieta deltoidea) base pairs, re-
spectively. Identical ITS sequences were detected
in D. lindenii and D. pulvinata subsp. pulvinata
(acc. nos. 38 and 80, respectively) and in D. sericea
(acc. no. 48), D. splendens (acc. no. 50), D. hem-
sleyana (acc. no. 83), and D. extensa (acc. no. 85).
The results of the distance analysis are shown in
Figure 2. Six major clades (I-VI as indicated on
Fig. 2) could be distinguished among the genus
Draba with relatively robust support for clades I-
Ш (Fig. 2)
We also indicated Schulz’s (1927) sectional clas-
sification (Table 2) in Roman numerals in the phy-
logenetic ITS tree (Fig. 2)
Fitch parsimony analysis with a subset of Draba
taxa from all major groups recognized in the overall
analysis resulted in 12 most ай وی trees
(MPTs) 208 steps in length, CI of 70% (autapo-
morphies lt ed RI of 69.99.
of the bootstrap and decay analysis are added to
The results
the strict consensus tree shown in Figure 3.
The overall analysis (Fig. 2), as well as the anal-
Volume 89, Number 1
2002
Koch & Al-Shehbaz 95
American Draba
Arabis pauciflora EUROPE
Arabis turrita EUROPE
pe; scabra
DRABA ARABOIDES USA
EROPHILA VERNA EUROPE
DRABA AIZOIDES EUROPE
Aubrieta deltoides EUROPE
rabis j va ed —
a EU
Arabis — USA
100
Arabis glabra U
Arabis lyalli USA
Arabis lignifera USA
Arabis divaricarpa USA
П 4
0.1 substitutions/site
Figure |. Neighbor-joining distance tree based on ITS
of Draba to other cruciferous taxa.
added to a previous analysis of tribe
Sequences. from
Arabideae (Koch «
"Arabideae core group? previously recogn
(L.) Bernh. herein we used the taxonomically more appropriate synonym Turritis glabra L.
below the branches from 1000 replicates.
ysis of a subset of taxa (Fig. 3), recognized the same
six major ITS clades among European and Ameri-
can Draba, including the two accessions of Ero-
European. Draba are confined to clade HI
2 and 3):
this clade is characterized by long internal genetic
phila.
together with European. Erophila (Figs. 2
distances separating European taxa from the re-
maining American Draba. The American species of
Draba are found in all remaining clades I. H, IV.
V, and \
The most basal clade, clade I, combines se-
Оа! аса vulgar IS
Cardamine flexuosa EUROPE
Р CUSIC
Halimolobus perplexa var. lemhiens
alimolobus perplexa var. ae pa
four taxa analyzed 1
ized by Koch et al (1999a). Distributional area is indicated.
Arabideae core group
EUROPE
Capsella rubella EUROPE
| Arabidopsis griffithiana ASIA
L Arabidopsi
abi — himalaica
abidopsis viret ASIA
Arabis glabra —
KIELLA ат Е сар USA
Arabis parishii USA
95- Arabis drummondii USA
sequence data to demonstrate the systematic relationships
ı this study (shown in capital letters) were
19994). Draba and Erophila were integrated into an
For Arabis glabra
Lal.
Bootstrap support is given
quences from mostly North American Draba taxa
D. cuneifolia, ace. no. 61; D. platycarpa, acc. no.
—
no. 72), but also one accession
no. 84). Clades I-III
are well supported by bootstrap values of 100%.
88%, and 93%, respectively. In contrast, clades IV.
V, and VI (combining the majority of taxa) are only
69: D. reptans, acc.
from Peru (D. araboides, acc.
weakly supported by bootstrap analysis (Fig. 2).
However, all clades were recognized in the parsi-
mony analysis with a subset of taxa from all six
clades. The only exception is D. scopulorum (acc.
Annals of the
Missouri Botanical Garden
67
Draba violacea (VIII) 86
Draba stylosa (?) 81
Draba Sedi n (VIII) 35
Draba depressa (V) 30
Draba pycnophylla (VIII) 46
Draba hammenii (?) 2
praos А. (XI) 32
raba funckiana a 87
denii (1X) 38
— lin
Draba sericea (?
Draba hemsleyana (VIII) 83
Draba macleanii (IV) 40
Draba splendens (VIII) 50
D soratensis (IV) 4
— Draba chionophila (ХІ) 8
— Draba nivicola (X) 58
- Draba alyssoides (
?) 23
— Draba pennell-hazenii (ХІ) З <@—
Draba extensa (V) 85
D nA na (X) 9
1 [—— Dra атса (ХІ) 73 ч —
80 Daba rositae (?) 14. аф
Draba зааг e 74
ӨГ Draba boyacana (?) 79
Draba arin F ree (?) 11
Draba pus
Draba чин М (8) 102
"T —
rV) 1
—— jaegeri
Draba pie Lim 34
Draba scopulorum (IV) 47 -«&—
Draba tomentosa (XIV) EU 113
raba tome
Draba ladina (XIV) 115 EU
Draba dubia
seh ramoisissima (XII) 76
6 ж
clade VI
clade V
clade IV
ntosa (XIV) EU 112
+-
(XIV) 1 114 EU
- 4-
— Draba funiculosa (11) 33
—[-
Р 59
Draba bellardii (ХІ) 7
—a L Praba cryophila (?) 5*
95 Draba cuatrecasana (?)1
ptans
?) 72
Draba ге
100 Draba platycarpa (XVI) 69
9 Draba cuneifolia (XVI) 61
94 Draba abor (XVI) 84
T^ uil av
— — (I) 10
94! Draba aizoides (1) 107 EU
| clade Il
| clade |
A:thriata rlalfnid, cil
| 7 Arabis alpina AFR
в
100
L Arabis alpina EU
pee Arabis bryoides EU
Arabis blepharophylla
| 0.02 substitutions/site
100
Arabis pumila EU
pe cias 111 ЕО clade Ill
110 EU
Volume 89, Number 1
2002
Koch & Al-Shehbaz 97
American Draba
Table 2.
parentheses in Figure 2.
Sectional classification system according to Schulz (1927
). Roman numerals correspond to those used in
Sectional classification
according to Schulz (1927)
Number of species and distribution
1. Aizopsis DC.
II. Linodraba O. E.
IH. Chrysodraba D(
IV. Rhabdodraba o. E. Schulz
. Schulz
28 spp..
Schulz | sp.,
57 spp.,
10 spp.,
8 spp..
MI ne. | sp.,
VII. Helicodraba O. E.
2 spp..
VIII. Cladodraba O. E. Schul 14 spp.,
IX. Dolichostylis (Turcz.) O.E. Schulz З spp.
X. Adenodraba O. E. Schulz
Mediterranean and rest of Europ
South American (Falkland e m m Patagonia)
Asia and North America
South America
South America (Ecuador and Peru)
NW Africa (Morocco)
NW Africa (Morocco)
South. America (Colombia south into Patagonia)
South America (Venezuela)
, Central America (Mexico and Guatemala) and South America
(Colombia, Ecuador, and Bolivia)
XI. Chamaegongyle O. E. Schulz 0 spp..
XIL Phyllodraba O. E. Schulz 29 spp..
ХШ. Nesodraba Greene | sp
XIV. Le е DC. 59 spp..
V.
XV. Drabella 23 spp..
XVI. Schulz
XVII. Schulz
ipie (Ral) O. E
Abdra (Greene) O. E.
© spp..
l sp.,
Europe,
South America (Colombia and Venezuela)
North
\me rica, К. Asia, and Himalayas
Asia. North America, and South America (2 spp.)
Asia, North America, and South pues (2 spp.)
North America and South. America (2 spp.
North America
no. 47). This species was positioned between clades
IV and II by distance analysis (Fig. 2) with higher
sequence similarity to taxa from clade IV, but par-
simony analysis positioned it in clade IV with high
confidence (Fig. 3). We obtained no significant phy-
logenetic structuring among these six major clades,
and the trees remained unresolved in this respect.
TRNL DATA
We considered 76 sequences of which 69 are
novel reports herein. In total we considered 72 dif-
ferent taxa. The alignment of the tral. intron is 382
bp in length and. contained 6 bp of the first exon
of the trnL(UAA) gene at the 5'-end and 30 bp of
the second exon at its 3'-end. The alignment of the
trnL-F-intergenic spacer region is 404 bp in length
and contained 5 bp of the second exon of the
trnl(U AA) gene at its 5'-end and 15 bp of the trnF
gene at the 3'-end. Both alignments were combined
to a final alignment 786 bp in length consisting of
the entire tral. intron as well as the second exon of
the trnl. gene and the complete tri L-F-intergenic
spacer. A region of six base pairs between the in-
tron and the spacer that showed no variation among
Draba accessions sequenced by Widmer and Bal-
tisberger (1999p) is missing. The whole data matrix
is interspersed with 24 gaps (7 within the tral, in-
tron, including 4 autapomorphies, and 17 within
the spacer region, including 8 autapomorphies, re-
spectively). Within the tral. intron 47 variable nu-
cleotide positions were detected (including 20 au-
tapomorphies); within the trnL-F spacer 52 out of
87 variable nucleotide positions were uninforma-
tive. The distance tree is shown Figure 4 with
support indicated by bootstrap greater than 50%.
Fitch parsimony analysis with a subset of taxa
similar to the ITS analysis resulted 82 MPTs
with 112 steps CI of 72%. and RI
79%. This subset comprises the same set of taxa
|
» analysis with additional Draba rositae
length,
used for ITS
(acc. no. 14), because this taxon has been separated
from other Draba taxa by a relatively high bootstrap
support of 63% (Fig. 4).
bootstrap and
Significant results of the
decay analysis are shown along
¢
Figure 2.
\rrows indicate positions of taxa that differ from the plastid analys
sligale concerte d evolution (for de ail. refer to text). Branch length of
S
used for triplet comparisons of ITS types to inv
Draba funiculosa (ace. no. ЗЭ. clade HH) has
Р
ш
correspond to those in
were not recognized by Se 'hulz (1927) are marked w
not been drawn to s
ible 1. In parentheses the есп — according to Schulz (19
ith °
Neighbor-joining distance tree of ITS sequence Наш: Bootstrap support is given from 1000 replicates.
) | PI
sis (see Fig. |). terisks mark accessions that were
accession са TS
that
a
ces of outgroup taxa а from irabis and brisa
HOR
e (refer to text). Taxonomic
27) is given. Ta
ITS sequer
have been published previously by Koch et al. (19994). Ael an Draba are indictaed by
98 Annals of the
Missouri Botanical Garden
99% == DRABA PULVINATA 80
=" clade VI
+4 ا DRABA BOYACANA 79
DRABA RAMOSISSIMA 76
83%
+2 clade V
DRABA HITCHCOCKII 64
53%
+1 DRABA OLIGOSPERMA 68
0,
ere DRABA SCOPULORUM 47 clade IV
DRABA PAYSONII 71
DRABA AIZOIDES 103
829
n EROPHILA VERNA 110 clade Ill
68% DRABA TOMENTOSA 112
+1
DRABA CUATRECASANA 1
63%
+1 clade Il
DRABA CRYOPHILA 5
: DRABA ARABOIDES 84
>5 clade
DRABA CUNEIFOLIA 61
ARABIS PUMILA
Figure 3. Strict consensus tree from o h parsimony using a subset of ITS sequences. Bootstrap support from 1000
replicates is seen above branches: dex alues are indicated below branches. Accession numbers following taxa
комны to “Table 1. Clades I-V Dee is to ITS clades l-V from Figure 2
branches in the strict consensus tree (Fig. 5). The
results of this analysis are congruent with the find-
ings of the overall distance analysis (Fig. 4). Five
major IN could be distinguished (clades A—E,
Figs. 4 and 5) with bootstrap support greater than
50%. There is no support to combine D. oligosper-
=
ma (acc. no. 68) and D. scopulorum (acc. no. 47
to any of these clades. However, D. oligosperma is
more closely related to cp genome types € and D
(Figs. 4
simony analysis.
The occurrence of these particular cp genome
and 5) as revealed from distance and par-
types according E their geographic origin is sum-
marized in Table 3
4) rec-
The more inclusive distance analysis (Fig.
ognized five major clades among Draba, with D.
cuneifolia (acc. no. 61), D. platycarpa (acc. no. 69),
D. reptans (acc. no. 72), and D. araboides (acc. no.
84) in clade A only distantly related to the remain-
ing clades. Plastome types from European Draba
were found mostly in clades В and D, but one ad-
ditional plastome type from Draba aizoides (acc. no.
108) (Widmer & Baltisberger, 1999b) showed an
intermediate position between clades E and C/D.
In the present analysis, three plastome types of D.
aizoides from Widmer and Baltisberger (1999b)
were included and all three were diverse. Draba
aizoides herein investigated from Germany (acc. no.
103) is more closely related to D. aizoides (acc. no.
107 from Widmer & Baltisberger, 1999b); it assorts
—
Volume 89, Number 1
2002
Koch & Al-Shehbaz
99
American Draba
raba dr. кш
— Draba
| саа pulvin
I Dra
— Draba solitara 6
Draba barclayana11
ana13
aa — berryi 18
[Draba schusten 15
ri,
de uds ТРЕТИ 33 —
Draba cruciata a.
89
Draba — 45
Draba boyacana 7
J —- Draba soratensis 49
Draba standley
Draba hidalgensis АЫ
Нот
чш clade Е
Draba araboides 8
L ..
ceca АП
Draba nivicola 58
—— Draba aureola 1
52 — Draba pulvinata 80
Es Draba aizoides 108 EU
RA
al aca 1105 EU
1 — aizoides 109 EU
a tomentosa 112 EU
мі 114 EU clade D
51 74 Draba ladina 115 EU
raba stylaris 106 EU
Draba scopulorum 47
raba rositae 14 «GH
= 63 | Draba phoney piel 73 —
Draba subumbellat
rta lol ensis 56
Draba magellanica 41
Draba jaegen
D aba incerta 65 clade C
ramosissima 76
Draba gilliesii 34
Dr. — helleriana 54 —
Draba paysonii 71
Draba rei NR 51
57| Draba hitchcockii 6
Draba glabell
Draba — 31 ——
Draba cryophi
D aizoides 107 EU
Draba pennell-hazenii З «8&— clade B
b sana 1
Draba aizoides 103 EU
Draba aizoides 104 EU
[^ Draba reptans 72
n h р, 3 Ras
clade A
100 5 Draba araboides 84
Draba platycarpa 69
| 0.05 substitutions/site |
Figure 4. Neighbor-joining distance tree of trnL sequence data. Bootstrap support is given from 1000 replicates.
Arrows indicate positions of taxa that differ fi "ig. Accession numbers of taxa correspond
to Table 1. European Draba are indictaed by
rom n the ITS analysis (see
100
Annals of the
Missouri Botanical Garden
г DRABA PULVINATA 80
0,
5 clade Е
lo DRABA BOYACANA 79
DRABA OLIGOSPERMA 68
76% DRABA TOMENTOSA 112
— $—1l1 dau
DRABA STYLARIS 106
0
52%
+1 o DRABA MAGELLANICA 41
13 h |
DRABA JORULLENSIS 56
0
100% 66% DRABA ROSITAE 14
>5 == clade С
+
DRABA HITCHCOCKII 64
Р DRABA RAMOSISSIMA 76
af / |
DRABA PAYSONI 71
DRABA SCOPULORUM 47
DRABA CUATRECASANA 1
72% ا DRABA AIZOIDES 103 clade B
+2
DRABA CRYOPHILA 5
os DRABA CUNEIFOLIA 61
clade A
Ll DRABA ARABOIDES 84
Figure 5. Strict consensus tree from Fitch parsimony using a subset of trnl sequences. Bootstrap support from
1000 replic ales is seen above branches; dec cay
values are indicated below branches
. Accession numbers following taxa
correspond to Table 1. Clades A-E correspond to trnL clades A-E from Figure 4
within clade B (Fig. 4).
aizoides are separated by seven — Some
7. D. oli-
gosperma, асс. no. 68, and D. aizoides, ace. no. 107
from Widmer & Baltisberger, 1999b) are not sup-
ported significantly within any of the five clades A
to E (Figs. 4
oligosperma is positioned basal to clades C and D
These two plastomes of D.
plastome types (D. scopulorum, acc. no. 4
and 5). The plastome type of Draba
in distance and parsimony analyses (Figs. 4 and 5).
The plastome type of D. scopulorum is supported
within clade D by distance analysis (Fig. 4); how-
ever, this is not supported by a bootstrap value
greater than 50%, and parsimony analysis with a
reduced data set (Fig.
at all. cpDNA sequence data from Erophila were
5) provided no information
not available, and research of the authors focusing
on annual Eurasian Draba is in progress.
ITS VERSUS frnl DATA
There is some congruence between phylogenetic
trees from both data sets (Figs. 2 and 4): (1) Amer-
ican D. cuneifolia (acc. no. 61). D. platycarpa (acc.
no. 69), D. reptans (acc. no. 72), and D. araboides
(acc. no. 84) are separated from all remaining Dra-
ba taxa, American as well as European material,
grouped in ITS clade I or corresponding
trnL clade
and are
A; (2) the majority of American Draba
is combined to ITS clade VI, which largely corre-
sponds to trnL clade E; (3) taxa that were integrated
Volume 89, Number 1
2002
Koch & Al-Shehbaz
American Draba
Table 3. Distribution of plastome types A-E among
Draba accessions analyzed herein according to their sam-
pling area. Unic “ambiguous”
jue or plastome type of D.
scopulorum, D. oligosperma, and D. aizoides (acc. nos. 47.
68. and 108, respective
bers of accessions showing a particular plastome type are
Чу) have been excluded here. Num-
indicated.
= : Plastome type
Origin of y 1.9
(according to Fig. 2)
accessions A B C D E
Europe 3 0
LS 3 8 5
Peru 1 8
Venezuela l 6
Colombia 3 l 5
one 2 l
hile 1 l
ae 2 3
Bolivia 2
Ecuador
into ITS clades IV and V possess a plastome type
typical for taxa from clade C; and (4) European
Draba is divided into two groups among both data
sets (clade B and D in the plastid data set: and
significant subgrouping within clade HI in the ITS
data set including Erophila), but the separation of
taxa forming clades B and D does not correspond
to that in the ITS clade HI.
However. eight accessions (D. pennell-hazenit,
no. 14: D. farsetioides,
33; D.
helleriana, acc. no.
acc. no. 3: D. rositae, acc.
D. funiculosa, acc. no.
47; D.
oligosperma, acc.
no. 73)
when both phylogenies (Figs. 2 and 4) were com-
acc. no. 31; scopu-
54 D.
no. 68: and D. streptocarpa, acc.
lorum. acc. no.
from America assign to different clades
pared with each other. In addition. plastome types
from Draba aizoides are found at three different po-
sitions (Fig. 4). Most American Draba species with
ITS type VI (Fig. 2) possess a plastome type group-
E (Fig. 4). However, American Draba
14). no. 54).
and D. streptocarpa (acc. no. 73) from ITS clade VI
ing into clade
rositae (acc. no. D. helleriana (acc.
do have the plastome type C and not type К, as in
the majority of taxa investigated, and within D. pen-
nell-hazenii (ace. no. 3) and D. farsetioides (acc. no.
31) from ITS
from clade B. Draba oligosperma (acc. no. 68) from
clade VI we found a plastome type
ITS clade IV (Fig. 2) and D. scopulorum (acc. no.
47). which is also grouped to ITS clade IV in par-
simony analysis (Fig. 3). do not have the plastome
type € (Fig.
clade IV. However, both plastomes of these taxa, D.
4) as do the remaining taxa of ITS
scopulorum and D. oligosperma, are found periph-
eral to plastome clade C (Fig. 4). Interestingly,
American D. funiculosa (ace. no. 33), with an ITS
type closer to European Draba, showed a plastome
type grouping to clade E (Fig. 4).
DISCUSSION
STATUS OF TRIBE ALYSSEAE
PHYLOGENETIC POSITION OF
AND THE
GENUS DRABA
Preliminary molecular phylogenetic analysis
—
Koch, unpublished) suggests that the morphologi-
cally poorly defined tribe Alysseae is not. mono-
Two of Schulzs (1936) tribes. Lunaricae
and Drabeae. Janchen (1942)
and Hayek (1911) as subtribes of the Alysseae. The
Al-Shehbaz
(1987) in the Alysseae, was shown by Zunk et al.
phyletic.
were recognized by
genus Camelina, which was treated by
(1999) to be very close to genus Capsella, a genus
traditionally placed in the tribe Lepidieae. A ma-
jority of the 650 species of the Alysseae (sensu
Hayek, and
genera Alyssum (170 spp.) and Draba (350 spp.).
—
Janchen. Al-Shehbaz) belong to the
Fhe present analysis supports a closer relationship
among Draba, Eurasian Arabis. and the European
Aubrieta. Eurasian Arabis and Aubrieta have been
shown to represent a core group of the tribe Ara-
bideae (Koch et al.. 1999a). However, as tradition-
ally delimited by Janchen (1942). Hayek (19101).
Schulz (1936), the tribe Arabideae breaks down as
1999a) (Fig. 1).
Similarily. the tribe Lepideae is not monophyletic.
an unnatural group (Koch et al..
and several of its constituent genera are sister laxa
in other tribes of the Brassicaceae (Koch et al..
2000, 2001).
from all molecular studies of the
that the tribal relationships of Schulz (1936).
Janchen (1942). Hayek (1911).
based solely on morphological characters, are high-
А | 8 g
The consistent conclusion reache d
Brassicaceae
S
and which are
ly artificial and do not reflect phylogenetic rela-
tionships. Such artificiality was also elucidated at
the generic level, especially by molecular studies
on Cochlearia (Koch et al., 1999b), Thlaspi (Mum-
menhoff et al.. 1997a, b), and Arabis (Koch et al..
1999a, 2000, 2001). Our findings should encourage
other investigators of cruciferous genera to test phy-
logenetic assumptions at a higher taxonomic level.
HYBRID ITS TYPES AND RETICULATION
We found no ambiguous nucleotide positions
when sequencing Draba ITS regions, although hy-
brid origin for several taxa is likely because of
some characteristics frequently found in species
groups showing hybridization and introgression: (1)
incongruencies between nuclear- and plastid-de-
102
Annals of the
Missouri Botanical Garden
Table 4
References for chromosome numbers among Draba taxa under study. For taxa that were recognized by
Schulz (1927) information about his sectional treatment is given (see also Table 2).
Sectional
classification
according to
Schulz (1927) Taxon Chromosome no. and reference Ploidy level
Not mentioned D. eryophila 2n = 48 (Galland & Pfitsch, 1986b) hexaploid
by Schulz
Adenodraba D. jorullensis n — 12 (Beaman et al., 1962) triploid
Aizopsis D. aizoides 2n — l6 (Hess et al., 1977) diploid
Calodraba D. gilliesii n — 24 (Boecher, 1966) hexaploid
Chamaegongyle D. bellardii 2n = 48 (Galland & Pfitsch, 1986a) hexaploid
D. chionophila 2n = 48 (Galland & Pfitsch, 1986a) hexaploic
Chrysodraba D. oligosperma 2n = 32, 64 oe 1972) tetraploid, octoploid
D. paysonii 2n — 42 (Mulligan, 197 — *2
Doliostylis D. pulvinata 2n = 48 (Galland & Pfitsch, 1986a) hexaploic
Drabella D. crassifolia 2n = 40 (Price, 1979 НИР
Leucodraba D. carinthiaca 2n = 16 (Hess et al., 1977) diploic
D. dubia 2n = 16 (Hess et al., 1977) —
D. fladniziensis, 2n = | (Hess et al., 1977) dipl
D. glabella n= 40 (Mulligan, 1970) oc E decaploid
D. incerta п = B (Mulligan, 1966) 14-ploi
D. ladina 2n = (Hess et al., 1977) tetraploid
D. magellanica n = 32 (Heilborn, 1941) octoploid
D. stylaris 2n = M (Hess et al., s 7) tetraploid
D. tomentosa 2n = 5, (Hess et al. 7) iploid
Phyllodraba D. helleriana n = 8 (Ward, 1983), n = 9 (Ward & Spellenberg. 1988) diploid
D. mogollonica n = 16 (Ward, 1983) tetraploid
D. ramosissima n = 8 (N 1969) diploid
D. streptoc arpa n — 20 (Price, 1979) pentaploid
Rhabdodraba D. pickerengi 2n = 24 (Favarger & Huynh, 1965) triploid
Tomostima D. cuneifolia 2n = A (Rollins & Ruedenberg, 1971) diploid
D. platycarpa 2n = 32 (Rollins « a 1971) tetraploid
D. reptans 2n = © (Live & Love 2) diploid, tetraploid
2n = 32 (Smith, 1965)
rived phylogenetic hypotheses, (2) the occurrence
of polyploids (Table 2). Additionally, for European
Draba extensive hybridization and reticulation have
been shown previously (Widmer & Baltisberger,
1999a, b). Therefore, we have to assume that con-
certed evolution of putative ancestral ITS types re-
sulted in one dominant ITS copy. Different modes
of concerted evolution of ITS regions have been
discussed, e.g., for Gossypium (Wendel et al.,
1995a, b), roses (Vissemann, 2000), Quercus (Muir
et al., 2001), and other plants (Buckler et al.,
1997), and for some cruciferous taxa (Hilliella,
Cochleariella in Koch & Al-Shehbaz, 2000; Arad-
idopsis in O'Kane et al., 1996; Cardamine in Franz-
1998; Microthlaspi in Mummenhoff et al.,
Sequence divergence values of ITS types
ke et al.,
1997b).
from putative parents, which produced hybrids in
which concerted evolution has been documented,
7.8% (Koch «€ Al-Shehbaz, 2000:
270). Sequence divergence values for Draba ITS
range from 3.1 to
types exceeded 9%, and these might indicate a rel-
atively old age for the genus. Divergence time es-
timates of cruciferous plants are found in Koch et
al. (2000) for chalcone synthase and alcoholdehy-
drogenase and in Koch et al. (2001) for chalcone
synthase and maturase К. A comparison of these
time estimates with an ITS-derived phylogeny com-
19994)
shows that 1% ITS sequence divergence corre-
prising a similar set of species (Koch et al.,
spond to approximately 0.5 to 1.0 million years.
American Draba occur at different phylogenetic
positions, sometimes together with European Draba
(Fig. 4).
polyploids (Table 4), and it is likely that hybridiza-
plastome types They include numerous
tion and polyploidization played a major role in
their evolution. In order to test this assumption and
show interrelationships between different ITS types,
several comparisons of three ITS sequences from
three different ITS clades (Fig. 2)
alyze the distribution of variable nucleotides. These
were done to an-
Volume 89, Number 1
2002
Koch & Al-Shehbaz 103
American Draba
American taxa compared as such are represented
in Figure 2. First, American Draba
cryophila (acc. no. 5, clade H) was compared with
D. hitchcockii (acc. no. 64) and D.
V and VI, respectively. The
by asterisks
—
tucumanensis
(acc. no. 51) of clades
distribution of nucleotide diversity revealed that D.
hitchcockii had 21 characters in common with D.
tucumanensis and only 5 characters with D. ery-
ophila (acc. no. 5). Of the eight nucleotide positions
where D. hitchcockii differed from D. eryophila and
D. tucumanensis, three were identical to positions
from D. jaegeri (acc. no. 66, also from clade V).
Synapomorphic characters were found within the
TS2 regions of D. tucumanensis (acc. no. 51) and
D. hitchcockii (ace. no. 64), but not within D. ery-
no. 5) and D. hitchcockii (acc. no. 64)
no. 66).
Different results were obtained upon comparison
57) of clade VI (Fig. 2)
51) and D. hitech-
In this
ophila (acc.
or D. jaegeri (acc.
—
of D. jorullensis (acc. no.
with D.
cockii/ D. jaegeri (accession nos. 04 and 66).
case. both D. hitchcockii and D. jaegeri shared five
characters with D. jorullensis (acc. no. 57) and sev-
en with D. tucumanensis (acc. no. 51). Three of
eight characters of D. hitchcockii (ace. no. 04).
which were not found in D. tucumanensts (acc. no.
tucumanensis (ace. no.
51) and D. jorullensis (ace. no. 57), were synapo-
morphic with those of D. jaegeri (acc. no. 66). Fur-
thermore, D. tucumanensis (acc. no. 51) contributed
to all variable nucleotide positions in the l1TS2 re-
no. 64)
whereas D. jorullensis
gion shared with either D. hitchcockii (acc.
r D. jaegeri (acc. no. 66).
fico: no. 57) contributed nucleotide positions only
within the ITS]
preted as the result of hybridization between taxa
region. These findings are inter-
of clades IV and VI, as represented by D. tucu-
manensis and D. jorullensis, respectively, which
produced the hybrid ITS types found in clade V as
represented by D. hitchcockii and D. jaegeri (Fig.
Four additional comparisons are made to test the
possible role of hybridization in the formation of
ITS types of clade VI (Fig. 2) that includes no Eu-
ropean taxa: (a) D. eryophila (асс. no. 5. clade I1).
D. jorullensis (acc. no. 57, clade VI), and D. tucu-
clade IV): (b) D. eryophila
nos. 5 Э
manensis (acc. no. OL,
and D. helleriana (acc. and basal in
clade VD. and D. tucumanensis (acc. no. 51.
IV): (e) D. tucumanensis and D. jorullensis (acc.
nos. 51 and 57) and D. helleriana (ace. no. 54): as
well as (d) D. eryophila, D. helleriana. and D. jo-
5. 54. and 57, respectively).
clade
rullensis (acc. nos.
The first two comparisons (first, among ace. nos.
5. 51.
thal accessions 54 and 57 had only 6 characters in
57. and second, among 5, 51, 54) revealed
common with accession number 5 of clade H but
23 characters with accession number 51 of clade
IV. The third comparison (acc. nos. 51, 51. and 57)
revealed that all synapomorphic characters com-
bined exclusively accession numbers 54 and 57.
This might indicate close relationships of taxa from
clade VI, which includes accession numbers 54
and 57, and a monophyletic origin. The fourth com-
parison (acc. nos. 5, 54, and 57) provides no ad-
ditional data toward the hybridization of taxa from
either clade Hl or IV as a source for ITS types р
in clade VI.
characters with accession number 5. but accession
Accession number 54 shared only :
number 54 shared 21 characters with accession
number 57. Therefore, based on ITS sequences, it
was not possible to verify the hypothesis that ITS
types from clade VI evolved via concerted evolution
after hybridization.
Comparison of the maternal plastid phylogeny
Fig. i
the ITS-derived phylogeny indicates that plastome
see taxa indicated by black arrows. ) with
types found in D. pennell-hazenii (acc. no. З) and
D. farsetioides (acc. no. 31) similar to those found
in European taxa (plastid clade B) are distributed
among accessions with an ITS type of clade VI. By
contrast. D. funiculosa (acc. no. 33) has a plastome
type found in clade E, but its ITS copy is related
to European ITS types found in ITS clade HI. Plas-
tome types from plastidic clade € (D. rositae. D.
helleriana, and D. streptocarpa or acc.
13. respectively) contributed to the genetic consti-
'S clade VI.
plastome type E, which is distributed among Amer-
ican Draba of the ITS clade VI.
plastome type from the European D. aizoides (acc.
no. 108, Fig.
of D. aizoides (acc.
nos. 14, 54.
tution of taxa found in TI Interestingly.
is similar to a
1) but not from the other accessions
107, 109, 103), which as-
sort to plastome clades B and D. Extensive intra-
nos.
specific cpDNA haplotype variation alpine D.
aizoides was demonstrated by Widmer and Baltis-
berger (1999b). Therefore, one could speculate that
ancient gene flow and chloroplast capture might
have resulted in the similar transmission of Euro-
pean plastome types into American Draba and in
the constitution of the whole complex of taxa with
an ITS type from clade VI. Plastome types found
in taxa from this group mostly belong to plastome
type E (only 5 out of 51 accessions with ITS type
VI have been shown to possess other. plastome
types than plastome type E) and are only weakly
differentiated from each other. There is no resolu-
tion within clade E in the trnb-derived phylogeny.
Taking both approaches into account, the com-
parison of ITS types and the distribution of plas-
tome types among different clades reveals strong
104
Annals of the
Missouri Botanical Garden
evidence for extensive reticulation during the evo-
lution of the American complex of Draba. However,
concerted evolution of ITS types has resulted in
homogeneous ITS copies within single taxa: no dif-
ficulties were encountered while obtaining the se-
quences by direct cycle sequencing of the purified
PCR products. The genetic source of plastome and
ITS type variation probably originated from Eur-
asian taxa, for which very complex evolutionary
scenarios have already been described. Subsequent
hybridization probably took place among different
ITS
resulted їп
groups of American Draba (e.g., taxa from
clade H with taxa from ITS clade IV
taxa from ITS clade V), and some taxa from the ITS
clade VI crossed back with taxa from clades V or
IV possessing plastomes of type С. Only taxa from
ITS clade I (corresponding to trnL clade
which are distributed in America, represent a ge-
netically separated lineage (Figs. 2, 4).
These hypotheses might be simplified; however,
they will become more complex when additional
individuals from a single taxon or one population
are analyzed, For example, extensive ongoing gene
flow has been reported within Draba from the Alps
and Scandinavia (Widmer & Baltisberger, 1999a,
1992a-d;
1993). Although the study of herbarium material is
b; Brochmann et al., Brochmann et al.,
greatly affected by undersampling. it is rather re-
markable that in 8 out of the 74 samples of Amer-
ican Draba (> 10%) analyzed, incongruencies be-
tween ITS- and trnl-derived phylogenies were
detected. One has to assume that much higher lev-
els of incongruencies will be observed in genetic
analyses at the populational level.
SECTIONAL CLASSIFICATION OF DRABA
Draba is generally recognized as a natural genus
of about 350 species distributed primarily in the
Arctic, subarctic, and alpine regions of the North-
ern Hemisphere, with about 65 species in South
America along the Andes from Colombia to Pata-
gonia (Al-Shehbaz, 1987). In the present study we
focused on American Draba, including a limited
number of species from Europe (8 species from 10
accessions). Eleven species grow in Mexico and
1984). More than 100
grow in North America and Greenland (Rollins,
Central America (Rollins.
1993), with the ranges of about 20 of these extend-
ing into the Arctic and subarctic Eurasia
Schulz (1927, 1936) divided Draba (excluding
Erophila) into 17 sections (Table 2) that have been
Fernald, 1934; Al-Sheh-
baz, 1987) to be highly artificial. Schulz defined his
considered by some (e.g.,
sections primarily on the presence or absence of
A), all of
stem leaves, flower color (white vs. yellow), the
presence or absence of median nectaries, and style
length. Four of Schulz’s sections are no longer rec-
ognized in Draba. Section Helicodraba O. E. Schulz
was shown by Hyam and Jury (1990) to belong to
the Southwest Asian Graellsia Boiss. Section Ne-
sodraba (Greene) N. Busch was reduced by Ber-
kutenko (1995) to Schivereckia Andrzeiowski ex
Raf.) О. E. Schulz and
Abdra (Greene) O. E. Schulz are believed to form
DC. Sections Tomostima
—
the well-defined genus Tomostima Raf. (Price & Al-
Shehbaz, unpublished). Species as assigned to the
remaining sections are still maintained in the genus
and are not currently questioned (Table 2). Section
Acrodraba O. E. Schulz (1 species, Draba oreadum
Maire) is restricted to northern Africa, whereas sec-
tions Rhabdodraba О. E. Schulz (10 species), Ad-
enodraba O. E. Schulz (9 species). Tylodraba O. E.
Dolichostylis (Turez.) О. E.
species), Calodraba O. E. Schulz (14 spe-
Schulz (8 species).
Schulz (3
cies), and Chamaegongyle O. E. Schulz (6 species)
are distributed in Central and South America. With
the exception of section Aizopsis DC. (28 species),
which is almost exclusively European, larger sec-
tions include the North American and Eurasian
Chrysodraba DC. (57 species), the North American
and East Asian Phyllodraba O. E. Schulz (29 spe-
cies), European and central Asian Leucodraba DC.
(59 spp.). and the Asian Drabella DC. (23 spp.).
The last included two South American species that
doubtfully belong there.
Four out of the 17 sections sensu Schulz (1927),
which comprise only five species (1.5% of the total
in Draba), were not included in the present study.
Sections Helicodraba and Nesodraba (4 species)
have already been combined under different genera
as outlined above. Material of the North American
section Abdra (1 species) and African section Ac-
rodraba (l species) were not available. Schulz's
sectional classification is given in Roman numerals
in parentheses in the phylogenetic ITS tree (Fig. 2).
The phylogenetic analysis of the ITS sequence
data (Fig. 2) shows that all sections exclusively dis-
tributed in South and Central America (Table 2,
Rhabodraba, Doli-
chostylis, Adenodraba, Chamaegongyle) fall within
clade VI.
Draba aureola (ace. no, 101) of section Phyllodraba
(North America, Asia), D. crassifolia (ace. no. 59)
North America),
and D. cruciata (ace. no. 60) of section Chrysodraba
sects. Tylodraba, Calodraba,
This clade also includes such taxa as
from section Drabella (Eurasia,
(Eurasia, North America). This suggests that South
and Central American Draba probably evolved
from other sections distributed in the Northern
Hemisphere. With our molecular data it is difficult
Volume 89, Number 1
2002
Koch & Al-Shehbaz
American Draba
South-
North migration events (from South America to the
to distinguish possible, relatively recent
North) from the earlier North-South migration route.
However, in this case some morphologically defined
wider distribution
species should have a much
range, which is not the case.
A few species from exclusively South and Cen-
tral American sections fall within other clades. For
example. Draba bellardii (acc. no. 7) of section
Chamaegongyle belongs to the ITS clade Il. where-
as both D.
Rhabdodraba and D. gilliesii (ace. no.
47) of section
34) of sec-
scopulorum (acc. no.
tion Calodraba belong to clade IV.
Some level of congruence was found when com-
paring the ITS clustering of European Draba with
the sectional classification of Schulz (1927. 1936).
Taxa comprising the ITS clade HI were assigned to
sections Alzopsis, e.g., D. aizoides, and Leucodraba.
e.g.. D. tomentosa. These sections are primarily
Eurasian. Interestingly, D. funiculosa (ace. no. 33)
falls within this ITS clade,
branch (more than twice longer than the maximum
but the extremely long
distance found. within the total ingroup) that sets it
off from the remaining European taxa indicates that
its ITS type has considerably diverged. This partic-
ular subantarctic species comprises the monotypic
section Linodraba.
A comparison of Schulz’s (1927, 1936) sectional
classification with the present molecular study re-
veals that section Zomostima (XVI, Table 2). which
corresponds to the ITS clade | (Fig. 2) and irni
clade A (Fig. 4), is the most clearly supported.
Schulz (1927, 1936) and all European taxonomic
treatments maintained Erophila as a distinct genus
that is closely related to Draba. However. the pre-
sent ITS data show that it is better integrated into
Draba. and these results support the view held by
most North American. botanists (e.g... Fernald.
1934; Al-Shehbaz. 1987: Rollins, 1993). The only
morphological difference between Erophila and
Draba is the presence of bifid petals in Erophila
rather than entire to deeply lobed ones in Draba
(Al-Shehbaz, 1987).
In conclusion, our findings do not provide much
support for Schulz’s (1927,
fication, although they suggest some species ten-
1936) sectional classi-
tatively correspond to a few of his sections. How-
ever, several taxa (D. pennell-hazenii, D. rositae. D.
farsetioides, D. funiculosa, D. scopulorum, D. hel-
leriana. D. oligosperma, and D. streptocarpa, acc.
4, 31. 33. 47, 54, 08,
which fall into different clades when the ITS- and
—
nos. 3, 73, respectively),
пі -derived phylogenies were compared, indicate
extensive cross relationships not only among taxa
of primarily South and Central American sections,
folia (diploid, acc.
but also among sections with other geographical
centers of distribution, e.g., North America, Europe
Eurasia), and the subantarctic. A case in point is
—
the South American Draba magellanica from Ar-
gentina (sect. Leucodraba, XIV). which has a chlo-
roplast type (clade С, Fig. 4) not found within the
remaining species of this section (e.g... D. dubia
within tral. clade D). Two other species from Ar-
33. and D. magel-
no. 41) have chloroplast types E and
gentina (D. funiculosa, acc. no.
lanica, acc.
C. respectively
Although one might argue that in order to
better
within
achieve a understanding of the sectional
Draba,
needed, especially from the sections not included
classification more analyses are
in the present study, it is safe to conclude that our
molecular data clearly show that the conventional
sectional classification of Draba as conceived by
Schulz (1927. 1936) is an artificial one.
CHROMOSOME AND
APOMIXIS
NUMBER, POLYPLOIDY.
Because the present study is based on herbarium
material, no chromosome counts were made from
the plants investigated. However, several chromo-
some counts are available from the literature (Table
base chro-
1). From this, it seems likely that the
mosome number for Draba is x = 8. This is also
the base chromosome number for sister European
genera Arabis and Aubrieta (Fig. 1). Chromosome
data demonstrate the existence of diploid, triploid.
hexaploid, octoploid, and
= 8 (Table 4). Higher
chromosome numbers (2n = 112) have been re-
tetraploid, pentaploid,
decaploid taxa based on x
ported for D. incerta (acc. no. 65) from North Amer-
ica (sect. Leucodraba).
The well-defined clade I (ITS
sponding clade A (trnL data), comprising D. cunei-
data). or corre-
no. 61). D. platycarpa (tetra-
69), D.
no. 72), and D. araboides (no count
ploid, ace. no. reptans (diploid and
tetraploid, ace
or кинин number, асс. no. 84), included dip-
loid and tetraploid species (Table 4). In summary.
XVI)
comprises mostly diploid species (as known so far).
this clade (well defined as sect. Tomostima,
and also represents an ancestral clade to all
maining Draba. This finding will need further in-
vestigation because section Tomostima is restricted
to North and South
taxa served as ancestors for Eur-
\merica, and it is hard to be-
lieve that these
asian Draba. The consequence is the recognition of
genus Tomostima (Price & Al-Shehbaz, unpub-
lished).
The ITS data indicate a hybrid origin for Draba
106
Annals of the
Missouri Botanical Garden
ramosissima (clade V, acc. no. 76, Fig. 2), a species
reported to be diploid (Table 4). As outlined before,
direct ITS sequence comparisons demonstrated that
ITS types from clade V
types of clades VI and IV. Therefore,
most likely evolved from
ancestral ITS
in D. ramosissima speciation via hybridization must
have taken place at the diploid level and only dip-
loid parental taxa could have served as progenitors.
Another diploid, D. helleriana (acc. no. 54) has an
—
ITS type corresponding to clade Ш (Figs. 2, 3. 6
C (Figs. 4—6).
incongrueney could be best explained by gene flow
and a plastome type to clade This
between different groups of taxa, thus indic ating
another hybrid speciation at the diploid level.
cept for D. cuneifolia and D. reptans from section
Tomostima and D. helleriana and D. ramosissima
from section Phyllodraba, counts for the remaining
American Draba species showed polyploidy (Table
4). By contrast, triploids (D. jorullensis, acc. no. 56;
D. pickerengii, acc. no. 44), pentaploids (D. cras-
59: D.
and an aneuploid (D. paysonil,
streplocarpa, acc. no. 13).
pentaploid + 2
no. 71) have been reported (Ta-
sifolia, acc. no.
chromosomes, acc.
ble 4), and these taxa grouped E plastidic clades
C and D or ITS
Apomixis has been bibi: in a few species
clades IV and V
such as Draba verna or D. oligosperma (Mulligan
& Findlay, 1970; Mulligan, 1971, 1972, 1976:
Price, 1980), and it is likely that it might be more
widespread among American Draba. Therefore, in
Draba apomixis might be an additional mechanism
that led to the recognition of several hundred, mor-
phologically defined taxa. The occurrence of apo-
mixis has also been documented for other Brassi-
caceae, and North American Arabis parallels the
1966; Rollins, 1983).
North American Arabis, from which many species
situation in Draba (Bócher,
were more appropriately transferred to Boechera
(Live & Lóve, 1975; Weber, 1982, 1989), is a
young taxon, presumably of Pleistocene origin. This
genus exhibits remarkable morphological. ecologi-
cal, and physiological diversification leading to the
recognition of more than 50 species, and most of
them show the trait apomixis.
As suggested by the low sequence divergence
'S clade VI
(Fig. 2) or the corresponding trnL clade E (Fig. 4)
values, American Draba either of the I'l
probably are relatively young taxa similar to Boech-
era. Comparable low levels of molecular variation
have been found separating the North and South
Moench, sensu Meyer (1973,
r American Thlaspi L. sensu lato from their
1993: Koch et al.,
Europe several plas-
American Noccaea
1979), о
Eurasian relatives (Koch et al..
1998c). Contrary to this, in
tome types from D. aizoides were found by Widmer
and Baltisberger (1999b), some of them very close
to D. aizoides, accession number 107 (DA2, DA3,
DA9, DA6, DA8, Widmer & Baltisberger, 1999b),
separated by a maximum of seven mutations (in this
analysis corresponding to clade B, Fig. 5), and one
plastome type (ПАТ) was separated from Draba ai-
zoides, acc. no. 108, by two mutations. All samples
analyzed by Widmer and Baltisberger (1999b) orig-
inated from the Swiss Alps and demonstrate the
enormous plastome variation even within one taxon
(Draba aizoides) from a restricted area. ITS se-
quence variation among D. aizoides accessions (Fig.
2) is much lower, and all D. aizoides accessions
analyzed herein confined to one group within ITS
clade HI. In this group we also found опе D. ladina
(acc. no. 116); this is not unexpected, because this
taxon evolved by hybrid speciation involving D. ai-
zoides as one parental taxon (Widmer & Baltisber-
ger, 1999a)
The close relationships between the floras of
Central Asia and western North America are well
documented (Parks & Wendel,
es therein). The putative Bering bridge connected
1993, and referenc-
Asia and. North. America several times throughout
1987).
and this land bridge probably served as an immi-
the late Tertiary and Pleistocene (Parrish,
gration route for several Brassicaceae, including
Thlaspi (Payson, 1926) and Stroganowia Karelin &
Kirilov (Rollins, 1982). Palynological analyses from
the Venezuelan Andes document the occurrence of
Draba in Holocene deposits after late Pleistocene
deglaciation events (Salgado Labouriau et al.,
1988
Literature Cited
Acarkan, A., M. RoBberg, M. Koch & R. 'hmidt. 2000.
Comparative genome analysis reveals extensive conser-
vation of genome organisation for ibi thaliana
and Capsella ө = Plant J. 32: 55-62
Al-Shehbaz, I. A. . The genera " Ah ysseae (Cruci-
ferae; Brassicac ui in the Ead United States.
J. Arnold — — 185-244
Anderson, J. K . Wa 1999, Chromosome num-
bere — in ги tribe Brassiceae — aceae): Ev-
‚ Evol. 215: 255-
5
arwic ide
e" nce from isozyme number. Pl. Sy
— B. G., M. J. Sanderson, J. M. Po ^. Woj
ciechowski, C. S. Campbell & M. J. Donoghue: 1995.
The ITS region of nuclear ribosomal | valuable
ource of evidence — i ee аш eae e Mis-
souri Bot. Gard. 82: 247-277
Baum, D. A., K. J. ae sma & P och. 1994, A phy-
logenetic analysis of [NA — eae) a on
nuclear ribosomal DNA se quences. Syst. Bot. 19: 363—
388.
Beaman, J. H., ;. De Jong & W. P. Stoutamire. 1962.
C hromosome — in the alpine and subalpine baia
` Mexico е mala. Amer. J. Bot. 49: 41-5
ней nko, . 1995. Detective story about one Ls
Volume 89, Number 1
2002
Koch & Al-Shehbaz
American Draba
107
naean species of Cruciferae. Linzer Biol. Beitr. 27:
1115-1122.
йске W.. M. Huthmann & Н. Hurka. 2000. Evolution
of крз ЖЕ" {аха їп Е К. Вг. (Brassi
Den Nijs, К. Marhold & Н.
(e ditors), — RS in P s Habitats. Spec ial
e en s in d s and Biodiversity :
Bócher, T. W.
on — species. 2 Some
Biol. Skr. 14(7): 1-74.
Bremer, K. 1988. The limits of amino acid sequence data
in angiosperm phyloge netic reconstruction. Evolution
12: 7
2. d
сеае).
A
—
.E un rimental and cytologic E studies
Arctic and montane cruci-
fers.
der
Broc — 1902. Polyploid evolution in arc — alpine
Drab К aceae). Sommerfeltia Suppl. 34.
——. 1993
Reproductive strategies. of bio and
ipli pecu of arctic Drab (Brassicaceae).
P ist. E 185
ү R. Elven. 1992. g ہے
sequences of poly — in arctic Draba (Brassicaceae).
Evol. Trends Pl. 6(2 —
Borgen & I 1993.
ships aid chromosome — of N
of Draba (Brassicace
l. vol. 185: 55-
E cological and genetic con-
aba
Stedje. Crossing relation-
ordic populations
ith e — on the D. alpina
— Nordic J. E 13 : 121-
—— ——,. P S. Soltis & D. | Sola, 1992a.
gins of octoploid Se — endemic Draba cac-
uminum: Electrophoretic and morphological evidence
Nordic J. Bot. 12: 257-272.
———— & ——. |992b. Recurrent formation
and polyphyly ‘of Nordic polyploids in Draba (Brassi-
caceae). Amer. J. Bot. 79: 673-688.
YE "Soltis & P. S. Soltis. 1992c. P tic
аи and payee ny - —— polyploids
Draba (Brassicaceae). Pl. ‚ Ev t» 35-70.
. B. Stedje € L. D n. "1992 2d. Gene po acros
levels in Draba (Brassicaceae). Evol. '
0(2): ae E 4
Buckler.
и. of
Multiple ori-
e
Trends
P. Holtsford. 1997. The
— ка s and
phyloge netic ا ations. Genetics 145: 621-832.
М.
Baldwin & M.
Ippolito & '
Босана] DNA:
. Donoghue, B. €
ternal transcribed. spacers
m js congruence with morphology. Amer. J. Bot. 62:
903-918
Lo T, < K. L. Huynh. 1965. In: A. Live & О.
Solbrig (e г), IOPB chromosome number reports IV.
Taxon 14: 86-92.
Felsenstein, J. 1985. Confidence limits on phylogenies:
i approach using the bootstrap. Evolution 39: 78:
91.
1995.
version —
Fernald, L. 1934. Draba in te mperate — astern
America ` Rhodora 36: 241—261. 285-305, 314-344,
—371, 392—404. pls. 290-310.
Bleeker, R. Kohrt «€ H.
A..
1998. Molecular systematics at Cardamine and
PHYLIP. Package
University of Washington, Seattle.
Phylogeny Inference
Franz *
Hurka.
allied genera (Brassicaceae): ITS and non-coding chlo-
roplast E A. Folia Geobot. Phytotax. 33: 2252 10.
«Ж. Phitseh. 1986a. In: A. Live € D.
буе (е rs C tibiis number reports. Taxon 35:
195-19
* ————. 1986h. In: A. Live «€ D.
tors), C "ocn cm number reports. Taxon 35: 610-013.
Galland,
Love (edi-
Gielly, L. € P. Taberlet. 1994. The use aaah da d "e
to resolve plant phylogenies: — oding versus ‚ве-
п es. Molec. Biol. Evol. 160-77 77.
Нат. Н. H. Har T H. Mes & J. M.
E 1994. Melee ular cd — noncoding re-
gions of the chloroplast ee in the (
related species. Curr. Genet. 558-500.
Havek, A. von. 1911. Enturf eines SC rucifer
phyloge netischer Grundlage. Већ. Bot. Ce ntralbl.
127-335.
ssulaceae and
en Systems —
ты. О. ve
isk. Bot.
E;
— р, ^ Ex
Some осоо numbers in Draba.
35: 141-142.
sdb &
Se — mer angrenzender Cds e.
Hirzel. 1977. Flora der
Birkhauser, Basel.
. Jury. 1990. A review of the genus
¡Gala e me — Draba L. section Не odraba
Schulz Brassicaceae. Bot. J. Linn. Soc. 102: 9-22.
94 Das System der Cruciferen. Oesterr.
-Shehbaz. 2000.
айс» a ihe C hine "se AM Wins aceae): Evidence
from plastid tral. intron and nuclear ITS D
data. Ann. Missouri Bot. Gard. 87: 246—272.
& H. Hurka. 1999, Isozvme ме in the poly-
ploid complex Mic — perfoliatum (L. lever:
Morphology а and онн ior Flo-
ra 194
— K ы nhoff € К. Zunk. 1993. Isoelektrische
Fokussi rung der Untereinheiten der Rubisco in 7hlas-
i (Brassicaceae): Weitere pore auf
а 4 rae s Керегі. 104: 371-381.
— Huthmann & Н. MER a. 1998a.
speciation and evolution in the PHI с a x Coch-
learia L. 51—400.
Mummenhoff & H. Hurka. The b. Molecular
biogeography * evolution of the о perfol-
Molecular system-
\ sequence
eine. Formen-
Ixozv mes,
(Brassicaceae). Bot. Acta 111:
iatum s. |. polyploid complex — aceae): Chloro-
de DN A and nuclear ribosomal DNA restriction site
varialion. Canad. J. Bot. m — v
— & —— . Systematics and e
— Ki of i tolerant Thlaspi caeru-
lescens in торе: Evide
nce from genetic stud-
ies based on isozyme analysis. Biochem. Syst. Ecol. 26:
823-838
. Bishop & T. Mitchell-Olds. 1999a. Molecular
sy semi s and « e ut is of Arabis and Arabidopsis. Pl.
Bi ol.
Western Et
Д — —————,. 1999. Molecular phyloge-
netics of Cochlearia L. and allied genera based on nu-
clear ribosomal ITS DNA sequence analysis contradict
— concepts of e ir жаны) relationships.
РІ. Evol. 216: 207
4 T. pow; Il-Olds. 2000. Compar-
ative evolutionary analysis of the chalconesynthase and
alcohol dehydrogenase from e and Arabis rel-
atives. Molec. Biol. Evol. 17: | 183—149&
; 2001. A systematics
of the |
hum ae: Evi * from coding plastome
matk and nuclear Chs sequences. Amer. J. Bot. 88:
534—5
Кош C. 1998. Infraspecific morphometric and nuclear
genome size — in polyploid sa utella ени
(Brassica aceae а), Amer. J. Bot. 85: 14
Kowalski. 5. P, T. Im Lan, K. A. I oon & A. H. Pat-
erson. и c omparative mapping of Arabidopsis thal-
iana and Brassica oleracea chromosomes re
“als
108
Annals of the
Missouri Botanical Garden
lands of conserved organization. Genetics 138: 499—
510.
agercrantz, U. 1998. Comparative mapping between Ar-
abidopsis thaliana and Brassica nigra indicates that
rassica genomes have evolved through extensive ge-
nome replication accompanied by chromosome fusions
не frequent rearrangements. Genetics 150: 1217—
Live, E < D. Lóve. 1975. Nome nclatural notes on Arctic
plants. Bot. Not. mus > -523.
— Ф — fas nmm B chromosome number re-
ports LXXVI. Tax : m 974—598.
Меуег, Е. К. 1973. C :onspectus der Thlaspi-Arten Europas,
Afrikas — Vorderasiens. Feddes Repert. 84: 449—470.
979. kritische. Revision der Thlaspi- Arten Eu-
ropas, Afrikas und Vorderasiens. |. Geschichte, Mór-
phologie y C horologie. Feddes Re pert. 90: 129-15
Muir, G., leming & €. Schlótterer. 2001. Three
divergent DN A clusters predate the species divergence
in Quercus pus a (Matt.) i ie 2 and — из rahui L.
Molec. Biol. es 18: 112-
Mulligan, G. ^" 19 0. Chromosome numbers of the family
J. Bot. 44: 309-319.
Cruciferae III. E
1970. NAT studies
and i бше allies in Canada and Alaska.
Bot. — 1437.
I71. Cytotaxonomic studies of Draba species of
pay-
of — glabe lla
Canad.
Canada e Alaska: Т ventosa, D. ruaxes, and D.
Canad. J. Bot. 49: 1455-1460
19
12. Cytotaxonomic studies of Draba species in
хопи.
Canada g5 Alaska: oligosperma and D. incerta.
ت г. 1763-17
. The genus hr in ( anada Alaska:
Key summary. Canad. J. Bot. 5 393
& J. NF indlay.
197 0, Sexual re — tion and
ея in the genus Draba. Canad. J. Bot. 48:
209-
— M., A. Franzke & М. Koch. 1997а. Molec-
ular data reveal convergence in fruit characters, tradi-
tionally used in the classifi ation of Thlaspi s.l. (Bras-
sicaceae)—Evidence from ITS-DNA sequences. Bot. J.
Linn. Soc. 125: 183-199.
— — UN 1997b. Molecular phyloge-
netics of Thlaspi s.l. ad on —
plast DNA restriction site q
internal transcribed spacers of nuclear hanima]
A. Canad. J. Bot. 75: 469—482.
LG 19 ).
(Brassicaceae) base
variation and sequences of
* X Note 's on the infraspecific taxonomy and
le af — of Draba ramosissima Desv. Castanea
34: 210—
O'Kane, s. s в. Schaal & 1. А. Al-Shehbaz.
The origin of ен suecica (Brassicaceae
dicated by nuclear rDNA sequences. Syst. Bot
1996.
as in-
Parks, ( & J. F. Wendel. 1993. Molecular divergence
between Asian and North American Liriodendron (Mag-
noliaceae) with — ations for interpretation of fossil
flora. Amer. J. Bot. 77: 1243-156.
dd
Parrish, J. T 1987. G lobal pn and prse 0-
climate of the late Cre E us and Early Tertiary. Pp.
51-73 in E. M. Friis, JC — & P. R. Crane
(editors), The Origins of е and their Biolog-
ical Consequences. Cambridge Univ. Press, Cambridge.
19
Payson, E. B. 1926. The genus > Thlaspi i in North America.
Iniv. . Wyoming Publ. Bol. |:
Petit, C., F. Bretagnolle € E dd
1995 Evolutionary
consequences of E hybrid zones in wild
spec ies TI E E
Price, R. A.
: Systematics and Geography.
— of Wisconsin, li
. 1980. Draba streptobrachia (Brassicaceae), a new
species pA Colorado. : — 32: 160-169
——. | Palmer & Al-Shehbaz. 1994. System-
atic ba cae ra of Arabidopsis : — ular and mor-
Meyerowitz
Cold Spring
om — a crassa Complex (Brassica-
M.Sc. thesis, Uni-
adison.
— al perspectiv 7-19
& C. R. Somerville editors), Arabidopsis.
oe bor Laboratory |
Ramsey, J. & D. W. $ Уе ый 1998; Pathways, mecha-
nisms, and rates > aie formation in flowering
lants. Ann. iun ‘ol. Syst. 29: 467-501.
Rie — L. к. Doe. 1999, Hybrid classifi-
ation: — ге genetic — based studies of ex-
perimental hybrids. Ecology 80: 361—370.
— ———, J. Whitton € — 1999. Hybrid zones
and the genetic architecture of a barrier to gene flow
)etween A — r species. Genetics 152: 713-727.
Rollins. R. . A new species of the Asiatic —
ae ia a ruc ice ») from North America and i
bioge — implications. Syst. Bot. 7: 214—220.
——— 1983. Interspecific hybridization and taxon uni-
Amer. J. Bot. 70: 625-
—
formity in кї (Cruciferae
. 1984. Draba (Cruc — in Mexico and Guate-
mala. — Gray Herb. 213: 1-10.
————. 1988. — — new genus of the Cruciferae.
J. Jap. Bor 63: €
— The [^ ruc ae rae of Continental North Amer-
ica. Sono Univ. Press, Stanfore
. Rüdenberg. 1971. С hromosome numbers of
Cruc ш . II. Contr. Gray Неф: 17-133.
— Lauran M. L., V. Rull, S hubert & S. Va-
lastro. 1988. The establishment r vegetation after late
Pleistocene de ‘glaciation in the paramo de miranda Ve-
— Andes. Rev. Palaeobot. Palynol. 55: 5-18
Schulz, ‚ 1927. Cruciferae—Draba et — In: A
E SON [ died, Pflanzenr. IV. 105 (Heft 89): 1-396.
Verlag von Wilhelm ке Leipzig.
‚ Engle a K. Prantl (ed-
. 17В: 227-658. Verlag
eipzig.
охоте numbers v — өз
vo
Smith, E. B. 1 Chrom
ering plants. I. Trans. Kansas Acad. Sci.
Soltis, D. E. & P. S. Soltis. 1993. Molec es des и те
dynamic nature * polyploidy. €. R. C. Crit. Rev. Pl.
Sci. 3-27:
é — t Recurrent formation
and genome evolution TREE 14: 348—352.
Swofford, D. L. 9. PAUP*. Phylogenetic Analysis Us-
ing Parsimony — Other Methods). Version 4.0b2.
Sinauer, Sunderland, Massachusetts
Taberlet, ielly, G. Pautou & J. — 1991. Uni-
versal primers "for amplification of three — re-
pm of chloroplast DNA. Pl. Molec. Biol. 10
1109.
Thompson, | R. Lumaret. 1992. The evolutionary
dynamics of pied — oe establishment
EE 7: 3
nd persistence. us R )2—3(
Tolmac hev, 939. Draba. In: d a Komarov & N. A.
Busch itor ш URSS 8: d 649, 650
Urbanska, . Hurka, Ё ‚В. Ne uffe r & K.
alo. or гин and evolution in
Cardamine (Brassicaceae) at Urnerboden, Central Swit-
Volume 89, Number 1
2002
Koch & Al-Shehbaz
American Draba
109
zerland: Biosystematic and molecular evidence. Pl.
Syst. . 204: 233-256.
Vissemann, V. 2000. Molekulargenetische und morpholo-
gisch-anatomische Untersuc hungen zur E volutio:
Genomzusammensetzung von
Caninae (Dc.) Ser. Bot. Jahrb. Syst. 12
n und
¡ES
N t
|
Walters, 1964. Draba and — = G. Tutin
el xu — Fl. Europaea 1:
Ward. 03. Chromosome counts fom New Mexico
and ec rn Colorado. Phytologia 54: —30
& R. Spelle — p 88. Chromosome counts of
anglospe m е New Mexico and adjacent areas. Phy-
hup 64: —:398.
Weber, A. po New names and Fog e
eM in the Rocky Mountain flora
3
prin-
Phytologia 51:
аљ" 7
‚ Additions to the flora of Colorado ХИ. Phy-
REN өт. d 20—
We nisl. J. F к & T. Seelanan. 1995a. Bidi-
rectional inte nine ‘us concerted evolution following allo-
ploid spec iation in colton, онур)
84.
Acad. Sei. U.S.A. 92: 280-2
Proc. Natl.
— & ——. 1995b.
DN. i sequence from gossypioides reveals an-
Molec. Phy-
An unusual ribosomal
cient, cryptic : inlerge nomic introgression,
logenet. Evol. 4: 298-313.
Vhite. T. J.. T. Burns, S. Lee & J. Taylor. 1990. Ampli-
fication and direct sequencing of fungal ribosomal RNA
genes for phylogenetics. Pp. 315-322 in M. A. Innis,
) кы & T. J. White (editors).
Academic Press, New
Widmer, A. & M. Ba 1990.
idence for allopolyploid speciation and a single origin
of the narrow endemic Draba ladina (Brassicaceae).
: 1282-1289.
Baltisberger. 1999}.
cpDNA variation in alpine А
ss aceae): — re oe and population
structure. Molec. Ecol. 1415
Zunk. K. K. ba 7 P ni Шогы 1999,
netic relationships in tribe Lepidieae (Brassicaceae)
^ Yor
iltisberger. ITS molecula ev-
Extensive intra-
Draba aizoides |
Phyloge-
based on chloroplast DNA restriction site variation.
Canad. J. Bot. 77: 1504-1512.
EVIDENCE FOR MOTH AND Peter Goldblatt? and John C. Manning?
BUTTERFLY POLLINATION IN
GLADIOLUS (IRIDACEAE-
CROCOIDEAE)'
ABSTRACT
Pollination strategies of Gladiolus, one of the largest genera of the monocot family Iridaceae, are unusually diverse
and include various bee species, foraging either for nectar or for pollen, passerine birds, long-proboscid flies a
for nectar, hopliine beetles that use the flowers primarily as sites fo r assembly, and Lepidoptera. Pollination by insects
f this order comprises two cse different sets of pollinators, night-flying moths (Noctuidae or Sphingidae) and
butterflies (evidently se one species of Satyridae). These lepidopteran-pollinated flowers have quite different floral
adaptations and both types are specialist systems, although moth flowers may be pollinated by a range of different
moths. In Gladiolus moth- -pollinated flowers are usually large, long- ا and pale- — or mottled dull — to
brown, are usually richly scented, often open fully only at night, and produce relatively concentrated nectar that is
sucrose-rich. Butte ily flowers, in contrast, are fully open during the day, close ДА ог ере at night, are
often bright crimson to scarlet, usually with prominent white splashes on the lower tepals, but are also large, have a
long tube, and produce quantities of relatively dilute nectar, either sucrose-rich or hexose-rich. Comparing the polli-
nation systems of related species, we infer that night-flying moth pollination arose 6 times in the genus, whereas
butterfly pollination arose 3 times in the 165 species of southern Africa, and that the two lepidopteran pollination
systems in Gladiolus arose quite independently of one another.
Key words: butterflies, floral ecology, Gladiolus, Iridaceae, moths, nectar, pollination systems.
=
Pollination in the approximately 35 genera of the (Goldblatt et al., 1998a). We have also presented
Iridaceae of sub-Saharan Africa, where over 1000 evidence for passerine bird pollination in Gladiolus
species of the family occur, is remarkably diverse. and reviewed pollination systems for the entire ge-
Across the continent pollination systems include nus (Goldblatt et al., 1999, 2001). By comparing
four orders of insects (Coleoptera, Diptera, Hyme- and contrasting the evidence for moth versus but-
noptera, Lepidoptera) and passerine birds, as well — terfly pollination in Gladiolus, we show just how
as facultative, and possibly obligate autogamy. Pol- different the two systems are that use lepidopteran
lination by insects of the orders Diptera, Hymenop- pollen vectors.
tera, and Lepidoptera is itself diverse, involving a Pollination by large butterflies, so far in Gladi-
range of families of each order and different sets of olus known to involve only the satyrid, Aeropetes
floral adaptations and associated rewards to polli- tulbaghia, has already been demonstrated for two
nators. Here we present evidence for two different species of Gladiolus and was inferred for several
pollination strategies utilizing Lepidoptera in the more because of the distinctive floral features pre-
large, predominantly African genus Gladiolus. sent in species pollinated by this butterfly (Johnson
Elsewhere we have documented passive pollination & Bond, 1994). These features include a large, un-
in the genus by large-bodied apid-anthophorine scented flower, bright red perianth often with white
bees (Goldblatt et al., 1998a), hopliine beetles splashes on the lower tepals, a narrow but fairly
(Goldblatt et al., 1998b), long-proboscid flies in the long perianth tube, and the presence of large quan-
families Nemestrinidae and Tabanidae with elon- tities of nectar (Johnson & Bond, 1994; Goldblatt
gate mouth parts exceeding 15 mm (Goldblatt € & Manning, 1998). Night-flying moth pollination
Manning, 1999, 2000), and an active pollination has been inferred for nine species of Gladiolus
system involving pollen-collecting female bees based on the presence of features commonly asso-
! Support for this study by grants 5408-95 and 5994-97 from the National Geographic Society is е, ас —
edged. We thank B.-E. van Wyk, Rand Afrikaans University, Johannesburg, for the nectar analyses: . Kaiser, Givaudar
Roure Research Ltd., Switze vl ind, for scent chemistry analy ses; and Colin Paterson-Jones for the uin of a moth
— — maculatus
3. rukoff Curator af African Botany, Missouri Botanical Garden, Р.О. Box 299, St. Louis, Missouri 63166,
U.S.A. ба йн tet org
*Compton Herbarium, National Botanical Institute, P. Bag X7, Claremont 7
ac.za
ANN. MISSOURI Bor. GARD. 89: 110—124. 2002.
TT:
5, South Africa. manning(e?nbict.nbi.
Volume 89, Number 1
2002
Goldblatt & Manni
ing 111
Gladiolus Lepidoptera Pollination
ciated with moth pollination, including a pale-col-
ored perianth, presence of fairly concentrated nec-
tar, and a strong, often distinctive scent with a clove
component, and in some species floral odor pro-
duced only at night (Goldblatt € Manning. 1998).
Field observations and investigation of floral fea-
tures allow us to expand our understanding of these
pollination systems. Moreover, by using information
about phylogenetic relationships in the genus and
comparing the pollination systems of related spe-
cies, we can infer the number of times that each
system evolved within the genus and suggest an-
cestral species, thus developing hypotheses about
the evolution of the floral traits associated with pol-
lination by different Lepidoptera.
MATERIALS AND METHODS
SPECIES EXAMINED
Observations on the floral and pollination biology
of Gladiolus were made during the years 1993 to
1999 in the field in southern Africa and in living
collections at the Missouri Botanical Garden, St.
Louis. and Kirstenbosch Botanic Gardens, Cape
Town. Together with our research on the systemat-
ies of the genus (Goldblatt & Manning. 1998; Man-
ning et al., 1999), we have identified 20 species
that have flowers likely to be pollinated by Lepi-
doptera from a total of 165 species of Gladiolus that
occur south of the Limpopo-Cunene River axis. At
least an additional 15 species of tropical Africa
(Goldblatt, 1996) have similar flowers and may be
assumed to be pollinated by Lepidoptera. Southern
African. species fall in five of the seven sections
recognized in the genus by Goldblatt and Manning
(1998: Blandus,
nearifolius, and Ophiolyza). Species in two addi-
sects. Hebea, Homoglossum, Li-
tional sections of the genus, sections Acidanthera
and Decoratus, in tropical Africa also exhibit fea-
tures consistent with pollination by Lepidoptera.
Candidates for Lepidopteran pollination were
recognized by two separate suites of characters. The
first is the classic syndrome for moth pollination
(Faegri & van der Pijl, 1979) in which pale flower
color and production of a strong. rich odor are as-
sociated with some features only exhibited at night.
In Gladiolus such flowers are usually large. have
an elongate floral tube, shortly exserted unilateral
stamens, and the strong floral odor is often pro-
duced or intensified at night. Pigmentation may be
cream or whitish, but several species have darkly
mottled to nearly uniformly dark brown perianths.
A second syndrome of characters, described in de-
tail by Johnson and Bond (1994), combines bright
red floral pigmentation, a large flag-like presenta-
tion, absence of floral odor, and often lower tepals
marked with white splashes (we use the term tepal
for perianth lobe here).
SEASONALITY,
AND FLORAL
FLORAL PHENOLOGY,
PRESENTATION
LONGEVITY,
Direct phenological observations are presented
on 18 southern African species of Gladiolus made
during the years 1993 to 1998 in the field (Table
) and in living collections at Kirstenbosch Botanic
—
Gardens. Cape Town. Observations include mode
and timing of anthesis (i.e., opening of individual
buds), anther dehiscence, expansion of stigmatic
lobes. followed by withering of the perianth. Data
on seasonality are taken from Goldblatt and Man-
ning (1998). Compatibility relationships were not
examined for the study. Plant vouchers (Table
are deposited at the Missouri Botanical Garden
Herbarium, St. Louis (MO), and/or the Compton
Cape Town (NBG).
Herbarium.
NECTAR ANALYSIS
Nectar volume measurements were taken from
unbagged flowers in the field, reflecting both rates
of secretion and depletion, and from plants main-
tained in the laboratory, and not visited by insects
—
effectively representing bagged flowers). Experi-
ence with Iridaceae has shown that nectar charac-
teristics gradually change in species retained 1
water for periods greater than 24 hours, the nectar
usually becoming more diluted (Goldblatt et al.,
1995). Nectar from flowers on cut stems maintained
in water in the laboratory was therefore sampled
18 hours.
were picked and nectar was withdrawn from the
within To collect. nectar whole flowers
base of the perianth tube with 3 wl capillary tubes
after separating the ovary from the perianth base
(perianth tubes are too elongate and curved to allow
nectar removal directly via the mouth of the tube
with a capillary tube). Detaching the perianth from
the top of the ovary causes minimal damage. and
contamination of nectar by fluid from broken tissue
is not significant given the large volumes of nectar
produced by flowers of species under investigation.
The percentage of sucrose equivalents in fresh nec-
laboratory on a
tar was measured in the field or
Bellingham and Stanley hand-held refractometer
(0-50%) from five or more individuals per popu-
lation, unless fewer individuals were available. Ad-
ditional nectar samples were dried on Whatman fil-
van Wyk,
Afrikaans University, Johannesburg.
ter paper no.
Rand
HPLC sugar analysis.
1 and analyzed by B.-E.
using
112
Annals of the
Missouri Botanical Garden
Table |
(other collectors). All study sites are in South Africa.
Study sites and voucher information for species studied. Vouchers are housed at MO (Goldbatt) or at NBG
fri
Species
Study site and voucher
GLADIOLUS SECTION BLANDUS
series Blandus
G. cardinalis Curt.
H. Wright
G. insolens Goldblatt & J. С. Manning
: — uae Oberm.
G. carmineus C.
empervir 7. J. Lewis
PRIUS SEC ‘TION LINEARIFOLIUS
series Linearifolius
G. emiliae L. Bolus
W Cape, near Swellendam, Mar.,
W Cape, near Bain’s Kloof, Jan., Goldblatt & — s.n. no voucher
W Cape, cliffs at Hermanus, Feb., 292
W Cape, Piketberg, Jar
W Cape, Potberg. Mar “Goldblatt 4 Manning 10176
E Cape, Tsitsikamma Mts.
Goldblatt 1
tt & poe 10166
‚ May. Manning s.n. no —
Manning 1057
W Cape, near Riviersonerend, Mar., Stayner s.n.
G. guthriei F. Bolus
G. J.
G. nerineoides Lewis
GLADIOLUS SECTION HEBEA
GLADIOLUS SECTION HOMOGLOSSUM
s Gra
С. ee. Coldbisi & J. €. M
G. maculatus Sweet
anning
W Cape, Bain's Kloof,
W Cape, Helderberg Reserve, Somerset West, Jan., Runnals 463
W Cape, Jonkershoek Mts., Jan.,
Mpumalanga, near Morgenzon, Oct.,
May, Manning 1080
Esterhuysen 32847
Goldblatt & Manning 10071
E Cape, near Grahamstown, Mar., Dold & Weeks s.n.
W Cape, Simonstown, June,
Manning s.n. no voucher
Devil's Peak, May, Manning s.n. no voucher
G. recurvus L.
series Tristis
G. hyalinus Jacq.
W Cape.
G. liliaceus Houtt. W Cape,
С. longicollis Baker
». tristis
GLADIOLUS SECTION OPHIOLYZA
sitiflor
T. Мыш
G. saundersii J. D. Hook.
Fairfield, Caledon, Sep.,
— Long Tom Pass, Feb., — Mad 9822
Cape, near Bredasdorp, Aug., Barker 284.
W Cape. Helderberg Reserve, Nov., Runnals 513
W Cape, Gydo Pass, Goldblatt E la 974ЗА
Lion’s Hea
, Barker
— s.n. no vouche
KwaZulu-Natal, Hillcrest, Jan., Goldblatt & Manning 9854
E Cape, Naude's Nek, Feb., Goldblatt & Manning 9550
Additional species with similar flowers, assumed to be adapted for — by Aeropetes or night-flying moths,
i G. J. Le
include G. acuminatus F. Bolus (putatively moths) and js —
field observations of the floral ecology of these species.
but we have no
insect visitors for G. albens, G.
is (putatively butterflies),
also lack ask uin of ir
carmineus, G. cruentus, б. hyalinus, С. insolens, and б. E я but those species are included above because we
=>
iave nectar and other data from study populations.
FRAGRANCE
Floral scent was noted with the human nose in
the field and in cultivated plants. Presence of
scents too weak to be discerned in the open air was
recorded after individual flowers were picked and
placed in clean, lidded glass jars and stored in a
warm place. The contents of each jar was sniffed
after a minimum of 60 minutes (Buchmann, 1983).
The site of scent production was examined by im-
mersing flowers in aqueous neutral red stain. Scent
chemistry was examined by R. Kaiser, Givaudan-
Roure Research Ltd., Switzerland, by gas chroma-
tography using a DB-Wax Capillary column (Kaiser,
)3). Scents were captured in capsules through
—
which air was drawn by a vacuum pump from a
small, lidded chamber containing open flowers.
POLLINATION MECHANISMS AND POLLEN LOAD
ANALYSES
Observations of insects on Gladiolus flowers in-
volved at least 5 hours total per species and in
some cases up to 20 hours total per species. For
species inferred to be pollinated by moths, an ad-
ditional 4 hours of observation was made per spe-
cies during daylight hours to determine whether di-
urnal might also occur.
Observations included mode of foraging and wheth-
er insects contacted anthers and stigmas during vis-
visits by insects
Volume 89, Number 1
002
Goldblatt & Mannin 113
9
Gladiolus Lepidoptera Pollination
its to flowers. Study populations always included at
least 20 individuals at evidently undisturbed field
sites. Insects observed to probe the floral tube or
to brush the anthers or stigmas were netted when
possible and, in the case of moths, then immobi-
lized in a jar using ethyl acetate fumes. To prevent
contamination of the body of an insect with pollen
carried by another in the same jar, the bodies of
insect specimens were isolated from each other by
wrapping them in tissue prior to pinning. Individual
butterflies were netted for measuring and observa-
tion of sites of pollen deposition and then released.
Body length and proboscis length of insects was
Body length
was measured from the base of the labrum to the
recorded from captured specimens.
tip of the abdomen. Mouth-part length was mea-
sured from the base of the labrum to the tip of the
proboscis. Night-flying moths are not easily cap-
tured simply because darkness makes them diffi-
cult to locate. Use of flashlights covered with trans-
lucent. red. cellophane paper for illumination
=
significantly assisted observation and netting of
moths. Captured moths were identified by Douglas
Kroon (Sasolburg, South Africa), and both moth and
butterfly voucher specimens were deposited with
the South African Museum, Cape Town.
Identification of pollen on insect bodies was done
by gently removing grains from the body surface
with a dissecting needle. The residue from needle
probes was collected on glass slides and mounted
in 1-2 drops of Calberla’s fluid (Ogden et a
1974). Pollen grains were identified microscopical-
ly by comparison with reference to pollen grain
preparations made from plants flowering at study
sites. Gladiolus pollen grains are recognized. by
their large size, monosulcate aperture with promi-
nent 2-banded operculum. and perforate-scabrate
exine (Goldblatt et al.. 1991).
RESULTS
SEASONALITY, FLORAL PHENOLOGY AND
LONGEVITY, AND FLORAL PRESENTATION (TABLE 2)
Flowering times in the Gladiolus species of
southern Africa that are pollinated by night-flying
moths show no obvious association with their geo-
graphic ranges in the summer- or winter-rainfall
zones of southern Africa (Goldblatt & Manning,
998). Flowering of the flora peaks in late spring
(September to November) in the winter-rainfall zone
ut in summer and early autumn (December to
March) in the summer-rainfall zone. This coincides
with the middle or end of the period of optimal
plant growth, during or soon after the main rainy
periods. In the summer-rainfall zone. G. longicollis
flowers early in the season, mostly October and No-
vember, sometimes producing a second flowering
flush in December-January. The only other species
of the region showing adaptation for moth pollina-
tion blooms in spring before the first rainfall of the
wet season. This is out of phase with the main flow-
ering period for Gladiolus in the summer-rainfall
zone (Goldblatt & Manning, 1998
Gladiolus hyalinus.
—
In the winter-rainfall zone,
G. liliaceus, and G. tristis follow the main flowering
pattern for the region, but G. albens and G. macu-
latus flower early in the season, in autumn or win-
ter, and G. recurvus flowers in late winter or early
spring. Flowering out of phase with the flowering
peak is also characteristic of the two moth polli-
nated species of section Linearifolius. G. emiliae
and G. guthriet, which flower mainly in April and
May, although vegetative growth in these two spe-
cies is delayed until the winter and. spring when
leaves are produced.
Flowering of butterfly-pollinated species of Glad-
iolus is always from late December to April. excep-
tionally May, and this is when the only butterfly so
far recorded on Gladiolus species, Aeropetes, is on
the wing (Johnson & Bond, 1994). This means that
for the butterfly-pollinated species of the winter-
rainfall zone, flowering occurs several months out
of phase with the main growth and flowering peak
for the region. As a consequence these species
must also have special ecological adaptations to
support the unusual flowering patterns. These may
be a specialized habitat, such as cliffs. stream
banks, or waterfalls (G. cardinalis. G. insolens. G.
sempervirens), or a growth pattern in which leaf pro-
duction occurs when conditions are suitable, later
in the season (G. carmineus, G. stefaniae), or both
(С. nerineoides, G. stokoet).
Population density appears to be moderately dif-
fuse, and plants form extended populations with
flowering individuals standing 1-3 m apart. Species
of specialized habitats such as Gladiolus tristis
(moist to marshy sites). G. cardinalis (waterfalls), G.
carmineus (coastal sandstone cliffs), G. sempervirens
(seeps and wet forest margins) may be locally com-
mon and grow in dense stands.
The pattern of flower buds opening on an inflo-
rescence is acropetal. In all species, a mature bud
expands in the early to mid morning. and the open
flower typically lasts four days in Gladiolus species
(Goldblatt et al., 1998a; Goldblatt & Manning.
1999). In moth-pollinated Gladiolus species flow-
ering lasts longer. Flowers of a cultivated sample
of С. tristis lasted five or six days. and flowers of
G. recurvus lasted five to nine days. Flowers usually
open one to two days apart, and hence there are
Annals of the
Missouri Botanical Garden
Table 2. Floral and phenological data for southern African Gladiolus species with flowers adapted for pollination
by Lepidoptera. Species are arranged taxonomically according to Goldblatt and Manning (1998). Perianth tube length
was recorded at study sites and may not represent the range for
out of phase with the peak flowering time for the rainfall zone.
the species. An asterisk (*) indicates species flowering
Flower Perianth Main flowering Rainfall
Species color tube (mm) Scent lime zone
GLADIOLUS SECTION BLANDUS
series Blandus
G. cardinalis red, white streaks 32-40 попе late Dec.-Feb. — winter*
n lower tepals
G. carmineus crimson-pink, white 30-35 none Feb.—Mar. winter*
streaks on lower tepals
G. insolens scarlet-red ca. 38 none Jan.—Feb. winter*
G. stefaniae red, white streaks 35-45 none Mar.—Apr. winter*
on lower tepals
G. sempervirens red, white splashes 25-42 none Mar.-May winter*
on lower tepals
GLADIOLUS SECTION LINEARIFOLIUS
ries Linearifolius
А emiliae densely brown speckled 32-45 strong fruity Mar.- Apr. winter*
>. guthriei dull purple-brown 20-27 sweet-clove Mar.—June winter*
G. nerineoides scarlet 25-31 none Jan.—Mar. winter*
G. stokoei rmine-red 30-35 none Mar. winter*
GLADIOLUS SEC TION HEBEA
series Permeabilis
G. acuminatus cream 16-22 sweet-floral Aug.—Sep. winter
G. robertso whit 28-44 sweet-clove Sep.-Oct. summer*
GLADIOL Us SECT ION HOMOGLOSSUM
series Gracilis
G. albens white-cream 45-60 acrid-metallic Mar.-May winter*
G. maculatus cream, heavily speckled 23-35 sweet-floral May-July winter*
brown to dull purple
G. recurvus cream to pale pink 27-36 sweet-clove July-Sep. winter
series Tristis
G. hyalinus cream speckled brown, 25-26 usually odorless Sep.-Nov. winter
purple or green weet-clove
G. liliaceus beige. buff or pale orange. 40—45 strong sweet- Aug.—Nov. winter
light mauve at night clove at night
". longic 'ollis subsp. cream, lightly speckled 100—110 strong sweet- Sep.-Nov. summer*
p> collis clove at night
— white to cream 85-110 strong sweet- Oct.—Dec. summer*
clove at night
G. tristis white to cream 40-60 strong sweet Sep.—Nov. winter
clove at night
pred US SECTION OPHIOLYZA
entus red, white on lower tepals ca. none Jan.-Feb. summer
A — red, white on lower tepals ~ none Feb.—Mar. summer
often two or more flowers open at any time on an
inflorescence of three or more flowers. Flowers of
moth-pollinated species are usually partly closed or
at least the tepals are flaccid during the day, and
The reverse pattern holds for butterfly-pollinated
species. At sunset, the tepals of most species partly
or fully close, the tepals then loosely enclosing the
exserted anthers and stigmas. During the day the
open fully at sunset when the tepals become firm tepals become fully expanded again.
Flowers of many Gladiolus species have been
found to exhibit mechanical protandry (Scott Elliot,
1891; Goldblatt et al., 1998a; Goldblatt & Man-
ning, 1999), and the species studied here conform
and fully extended. In species of Gladiolus series
Tristis of section Homoglossum scent is released at
the same time, although a faint odor may be de-
tected during the day.
Volume 89, Number 1
2002
Goldblatt & Mannin 115
g
Gladiolus Lepidoptera Pollination
to this pattern. The anthers dehisce longitudinally
one to four hours after the tepals first unfold. This
depends to some extent on ambient temperature
and humidity, and anthers dehisce later in wet, cool
conditions. Pollen grains are clumped together and
pollen remains in the anther thecae until removed
by an insect. The three style branches, the distal
adaxial surfaces of which comprise the stigmas, are
loosely held together for the first three (to five) days
that the flower is open and lie laxly over on the
dorsal surface of the anthers. On the last day of
anthesis, the style elongates and the style branches
diverge, arching outward beyond the anthers. At the
same time, the conduplicate margins of the distal
half of each style branch unfold,
moist, sticky stigmatic surfaces of the now spath-
exposing the
ulate style branches. Only then are the stigmas of
a flower accessible to pollen deposition, and pollen
adheres to these areas following hand-pollination.
Thus of the four to six days that a flower is open,
it typically has three to five days in an exclusively
male phase during which time pollen is usually re-
moved from the anthers by insects. Anthers can be
seen with the naked eye to lack pollen after three
or four days if flowers were actively visited. By the
time the stigma lobes unfold the flower is then in
an exclusively female phase that lasts for the final
one (or two) days that a flower is open. Mechanical
self-pollination cannot readily occur, even if pollen
remains in the anthers by the time the receptive
stigmatic areas are exposed because of the spatial
separation of the pollen-bearing anthers and the
stigmatic surfaces. Gladiolus carmineus, which we
include here as possibly pollinated by large but-
terflies although we lack pollinator observations, 15
an exception. The style divides opposite the base
of the anthers and the style branches are tangled
in the dehisced anthers; thus, selfing could easily
occur if there were no incompatibility system (un-
known at present). M the G. carmineus study site
pollen had not been removed from the anthers, and
the stigmatic surfaces become dusted with pollen
from anthers of the same flower.
Species of Gladiolus are medium-sized, corm-
bearing geophytes, typically 45-120 cm high (Fig.
LA—F).
produce a single, unbranched flowering stem an-
Species pollinated by Lepidoptera typically
nuallv. Flowering population is synchronous
and lasts two to four weeks, and inflorescences are
typically secund spikes with the flowers facing to
the side and with the floral tube in an ascending
position. In G. nerineoides the flowers are crowded
at the apex of the flowering stem in more or less
spiral arrangement and it is the entire inflores-
cence, rather than individual flowers, that makes
for a conspicuous display (Fig. 1B).
Flowers pollinated by moths are moderate in size
to large, depending on sectional affinity (Fig. 1C—
=
`). In species of Gladiolus sect. Homoglossum the
perianth tube is mostly 23—60 mm long. but excep-
tionally 85-110 mm in G. longicollis subsp. platyp-
etalus. The dorsal tepal, usually slightly larger than
the other five tepals, is 32—45 mm long, thus short-
er to about as long as the tube (Table 2). In Glad-
iolus sects. Hebea and Linearifolius flowers are
somewhat smaller, and G. acuminatus has a floral
tube 16-22 mm long and tepals 15-21 mm long.
Irrespective of sectional affinity, the perianth tube
is more or less cylindric but slightly wider toward
the apex, with the slender lower portion 1.5-2 mm
in diameter, and fairly straight or gently curved.
Flowers are zygomorphic and the slightly larger
dorsal tepal is typically inclined while the upper
lateral tepals are spread outward. The lower tepals.
usually slightly smaller than the upper three, are
The
style and stamens are unilateral and arch to lie
held loosely together and directed forward.
close to and just beneath the dorsal tepal and are
thus contained within the enclosed space formed
by the ascending tepals. The filaments are shortly
exserled from the tube, or in G. longicollis not or
barely exserted, and the anthers lie parallel to one
another with the lines of dehiscence facing toward
the. center of the flower and the lower tepals. In
many moth-pollinated Gladiolus flowers the tepals
are somewhat attenuate with the tips recurved, a
feature most exaggerated in G. acuminatus and б.
recurvus.
Moth-pollinated flowers are either shades of
white to cream or are lightly to densely mottled with
dull purple to brown (Table 2: Fig. IC—F). Each of
the three lower tepals may have a weakly contrast-
ing. darker median band, often collectively referred
lo as a nectar guide (Goldblatt & Manning, 1998).
The bases of the tepals and the distal part of the
tube together form a wide throat leading to the nar-
row, proximal part of the tube. In the sense of Fae-
eri and van der Pijl (1979), these are gullet flowers
but often with a particularly elongated floral tube.
A remarkable feature of Gladiolus liliaceus is the
color shift that occurs at sunset. As the tepals be-
come more fully extended the normally beige. light
brown or rusty colored tepals take on a yellowish
background hue with the brown pigment changing
to light blue-mauve. First reported by Henry An-
drews in 1798, this color change has intrigued bot-
anists ever since and is thought to be a direct ad-
aptation to moth pollination (Goldblatt & Manning.
1998). A less pronounced color change al sunset
116 Annals of the
Missouri Botanical Garden
—
Figure 1. Comparison of the flowers of southern African Gladiolus pollinated by the satyrid butterfly Aeropetes
tulbaghia (A, B) versus night-flying moths mostly of the families Noctuidae and Sphingidae (C-F), with longitudinal
ections of flowers of some species. —A. G. sempervirens (sect. Blandus). —B. G. nerineoides (sect. Linearifolius). —
С. G. emiliae (sect. Linearifolius). —D. G. maculatus (sect. Homoglossum). —E. G. liliaceus (sect. Homoglossum). —F.
G. guthriei (sect. Linearifolius). Scale bar 10 mm.
Volume 89, Number 1
2002
Goldblatt & Manning 117
Gladiolus Lepidoptera Pollination
also occurs in the related G. hyalinus, the tepals of
which become paler and more translucent as day-
light fades. The darkly mottled tepal coloration may
be an adaptation for camouflage. rendering the
flowers less visible to nectar or pollen thieves dur-
ing the day. Johnson (1995) has suggested that the
maroon pigmentation of the moth-pollinated flowers
of the orchid Monadenia ophrydea Lindl. likewise
represents camouflage. Moths are believed to locate
flowers of this species solely by their scent.
Flowers pollinated by the Aeropetes butterfly are
fairly large (Table 2), with the exception of Gladi-
olus nerineoides, and have a perianth tube mainly
30—45
slightly larger than the other five tepals, is mostly
mm long. and the dorsal tepal. usually
to about
35-65 mm long, thus somewhat longer
In G.
somewhat smaller flowers have a floral tube 25—31
twice as long as the tube. nerinevides the
mm long and tepals 19-22 mm long, thus shorter
than the tube. The perianth tube is more or less
cylindric but slightly wider toward the apex. with
the slender lower portion 1.5-2 mm in diameter.
and fairly straight or gently curved. Flowers are zy-
gomorphic (barely so in G. nerineoides) and the
larger dorsal tepal either more or less erect, or
lightly inclined (G. cardinalis). while the upper lat-
eral tepals spread outward. The lower tepals are
usually slightly smaller than the upper three and
The
style and stamens are unilateral and extend out-
are held loosely together and directed forward.
ward. and well exserted from the tube. Exceptional
G. nerineoides has subequal tepals spreading more
or less at right angles to the tube and the filaments
are included, while the anthers may be partly ex-
serted or entirely included.
Butterfly-pollinated flowers are shades of red. oi
exceptionally deep pink (Gladiolus carmineus). and
the lower tepals are often streaked with white. In
G. saundersii the lower tepals may be almost en-
tirely white with irregular reddish speckling (Table
2). The flowers of all the butterfly-pollinated spe-
cies lack detectable odor (Table 2). In the sense of
Faegri and van der Pijl (1979), these are flag flow-
ers but with a particularly elongated floral tube.
G. nerineoides, and б.
The flowers of G. insolens.
stokoet are uniformly colored.
NECTAR ANALYSIS (TABLE 3)
Nectar produced by moth-pollinated species of
Gladiolus is sucrose-dominant and ranges in con-
centration from a low of 19.5% (С. emiliae) to
36.4% (G. tristis) sucrose equivalents. Nectar quan-
tities, measured from unbagged flowers, range from
2.2 to 12.4 pl.
Nectar of butterfly flowers is markedly variable
in character, and ranges from sucrose-dominant (6.
cardinalis, G. stefaniae) to sucrose-rich (G. cruen-
tus, G. saundersii) or hexose-rich (G. insolens. G.
nerineoides) according to the defintion of Baker and
Baker (1983).
low concentrations, ranging from a low of 18-21%
These nectars also have relatively
sucrose equivalents in G. stefaniae to 26.8% in one
of two populations of G. nerineoides examined.
FRAGRANCE
Moth-pollinated species of Gladiolus. with one
exception, produce strong, and often rich sweet
odors (Table 4). In species of Gladiolus series Tristis
fragrance is produced at nightfall and is weak or
evidently absent during the day. Fragrances vary
considerably as perceived by the human nose.
though often appear to have a strong clove com-
ponent and thus resemble the scent of stocks (Ma-
thiola) or carnations (Dianthus). The odor produced
by G. albens (sect. Homoglossum) is strikingly dif-
ferent. and is somewhat acrid and metallic, while
the odor produced by G. emiliae (sect. Linearifolius)
is fruity, with elements of coconut and pineapple.
Scent production in б. hyalinus is evidently un-
common. In four populations we examined (Die
Gale. Gydo Pass, Lions Head, Nieuwoudtville). all
in the west of its geographic range, flowers pro-
duced no detectable odor, but collection notes with
some herbarium records from the eastern half of its
range indicate the presence of a strong, sweet scent.
Absence of odor is combined with mottled, brown-
ish or purplish coloration in these populations.
Scent compounds mostly belong to different
chemical classes of scents from those of bee-polli-
flowers (Goldblatt et al., 1998a),
for linalool, which is present in some bee-pollinated
nated excepting
species, notably G. alatus, while benzyl acetate is
also present in G. jonquiliodorus. Scent chemical
profiles differ considerably among moth-pollinated
species, even though linalool is a common compo-
nent (Table 4) and is the predominant compound
in G. maculatus. As in bee-pollinated species of the
genus. numerous compounds combine to produce
the characteristic scent of each species, and
many as 39 compounds were identified in G. re-
currus (R. Kaiser. pers. comm.). Surprisingly, G. ac-
uminatus. the only species of Gladiolus sect. Hebea
examined, shares no compound with the moth flow-
ers of species of section Homoglossum to which be-
long the other species analyzed. The spicy-clove
type scents of G. liliaceus and G. tristis appear to
derive from different compounds, eugenol in the
former and eucalyptol in the latter.
118
Annals of the
Missouri Botanical Garden
Table 3.
analyses were provided by B.-E.
Wyk, Ra
Available nectar characteristics of species of southern African Gladiolus pollinated by Lepidoptera. Nectar
nd Afrikaans University, Johannesburg, South Africa. Number of samples
(n) is the same for volume and concentration columns. Data marked with an asterisk (*) are from Johnson and Bond
(1994
ў % Nectar % Range of sugars Sugar ratio
Gladiolus volume pl concentration S/F + G
i (n) +SD) Fructose Glucose Sucrose (n)
GLADIOLUS SECTION BLANDUS
series Blandus
G. cardinalis *9.4 (12) *24.8 (n/a) 9 16 75 3.0 (1)*
G. carmineus 1.74.8 (10) 26.1 (3.6) 18 30 52 1.08 (1)*
G. insolens 4.8-7.6 (3) — 39 38 23 0.26 (1)
G. stefaniae 8.6-11.0 (2) 18-21 10-16 18-25 59-72 1.89 (2)
*10 12 78 3.5 (1)
GLADIOLUS SECTION HEBEA
series Permeabilis
G. robertsoniae (2) — — — — —
GLADIOLUS SECTION иш
series Linearifolius
G. emiliae 3.5-6.6 (3) 22.3 (1.5) — — —
5.1-6.4 (2) 19.5-20.5 8-15 5-9 76-87 4.4 (2)
G. guthriei 4.44.8 (5) 31.4 (3.0) — EM —
G. nerineoides Helderberg 1.84.2 (6) 26.8 (2.3) 12-26 23-36 37-63 1.85 (3)
ershoe 2.9-4.6 (3) 23.7 (2.6) 8 46 26 0.35 (1)
GLADIOLUS SECTION HOMOGLOSSUM
series Gracilis
G. albens 3.8-5.6 (4) 27.0 (2.2) — — — —
G. maculatus 4.8-6.0 (2) 28-30 — =e ==
r TECUTVUS 4.1-13.7 (6) 31.9 (2.0) 6-7 11-14 80-82 4.3 (2)
series Tristis
G. hyalinus Gydo Pass 2.2-4.7 (3) 35.8 (4.8) 2-3 7-8 90 9.0 (2)
Lion’s Head — — 0-4 2-7 89-98 13.8 (3)
G. liliaceus 3.5-6.4 (5) 35.2 (1.2) 1 2-3 36—97 26.0 (3)
G. longicollis 2.7-3.7 (3) 24.7 (2.5) 6-27 9-23 50-85 2.2 (2)
subsp. platypetalus 5.1-7.9 (3) 28.3 (1.5) 6-11 6-11 78-88 4.3 (2)
. tristis 8.5-12.4 (5) 36.4 (2.1) — — — —
GLADIOLUS SECTION OPHIOLYZA
G. cruentus 4.8-5 : 20.7 (2.5) 26. 39 35.5 0.55 (2)
G. saundersii 14.7-20.1 (5) 24. 15 2 43 0.75 (1)
POLLINATION MECHANISMS AND POLLEN LOAD
ANALYSES
Moth pollination. Observations of pollinators
on 7 species of the 11 putatively moth-pollinated
southern African Gladiolus species showed all of
them to be visited by night-flying moths (Table 5).
Visitors included sphinx moths alone (G. longicol-
lis), sphinx and noctuid moths (G. emiliae), or a
range of small and larger moths including Noctui-
dae and other families. No other animals were not-
ed visiting these species either during the day or
at night excepting for a single male Anthophora div-
ersipes (Apidae: Anthophorinae) captured while at-
tempting to forage for nectar on the long-tubed
flowers of С.
contact of both anthers and stigmatic surfaces, and
recurvus. Its activities did result in
Gladiolus pollen was recovered from its body. This
bee should probably not be considered a regular
pollinator of the species since the reward offered is
not accessible to the bee, which has a tongue up
to 12 mm long. while the tube of G. recurvus is at
least 27 mm long.
Details of moth visits are limited because of the
difficulty of observing their activity in the dark or
under low intensity red light. Except for species of
Sphingidae, moths settled on flowers, grasping the
lower tepals, before inserting their probosces into
the floral tube (Fig. 2). Too few visits were noted
on any species for us to make observations of the
duration of visits—we were determined first to cap-
ture visiting moths for identification of species and
location. of sites of pollen deposition. Captured
Volume 89, Number 1
2002
Goldblatt & Manning 119
Gladiolus Lepidoptera Pollination
ladiolus with flowers adapted for moth pollination (R. Kaiser, pers. comm.).
J
7
Scent characteristics of selected species of southern African €
Table 4.
Scent composition (% constituents above 2%)
)-
(E, E)-alpha y
(E)-
ocimene
epoxide
poxy-3,
+
F
Benzyl
cinnamic
alcohol
farnesene
and farnesol
7-dimentyl
(E)-
Phenylacet-
aldehyde
Methyl
J
ocimene | 6-octadiene
Eucalyptol
benzoate
Linolool
or benzoate Eugenol
\ alcohol
dehyde
description
Species
6.2
31.0
14.1
floral
G. acuminatus
4.0
6.9
lily
G. maculatus
9.5
69.5
12.6
gardenia
recurvus
G.
10.5
3 61.5
5.8
clove
G. tristis
moths were found to carry Gladiolus pollen on the
proboscis, but no pollen was recovered from moth
bodies. After being netted and transferred to a kill-
ing bottle, moths lose many of their body scales and
may have also lost any pollen they might have car-
ried. Nevertheless, those moths with elongate pro-
bosces, including the sphingids Hippotion and
Agrius and the noctuid Cucullia, must be regarded
as legitimate pollinators of the species on which
they were captured. Moths in the families Archiidae
and Geometridae with shorter probosces, 10 mm
long or less. that were also captured, cannot reach
nectar in the floral tubes and must be regarded as
accidental visitors, perhaps attracted by the strong
floral odors.
Butterfly pollination.
were seen to be visited by Aeropetes tulbaghia and
Five species (Table 5)
another four species (Gladiolus carmineus, б.
cruentus, G. insolens, and G. stokoei) have similar
flowers and are inferred to share this pollination
strategy. This large butterfly has a wingspan of ca.
80 mm. a large body ca. 20 mm long, and a pro-
boscis 28-34 mm long (Johnson & Bond, 1994;
Johnson. 1994). Aeropetes was seen fluttering
around. flowers of G. cardinalis, G. nerineoides, G.
saundersii. G. sempervirens, and G. stefaniae (span-
ning three sections, Blandus, Linearifolius, and
Ophiolyza. Table 5). sometimes settling for 25 to
130 seconds, and then to fly to other flowers of the
same species. As Johnson and Bond (1994) have
noted, the behavior of butterflies included inspec-
tion visits when individuals flutter above flowers
without settling or feeding. During such visits, the
ventral part of the thorax, abdomen, and wings may
brush against the well-exserted anthers or stigmatic
surfaces of species like G. cardinalis, G. sempervi-
rens, and G. stefaniae (sect. Blandus). Pollen and
stigmatic surfaces of G. nerineoides (sect. Lineari-
folius) and G. saundersii (sect. Ophiolyza) are only
contacted during feeding visits when insects settle.
grasping the tepals and inserting their proboscis
into the perianth tube. In G. nerineoides pollen is
deposited only on the upper proboscis because the
anthers are not or are only shortly exserted from
the floral tube. Pollen is deposited on the head.
antennae, and dorsal part of the thorax of G. saun-
dersii because the hooded dorsal tepal, under which
the anthers lie, forces a feeding insect to orient its
body nearly vertically as it grasps the upper lateral
or lower tepals. No other animals were ever seen
visiting those Gladiolus species that attracted Aero-
petes, and this single insect must be assumed to be
their sole pollinator.
The innate attraction that Aeropetes exhibits for
red flowers, particularly those of large size, or col-
120
Annals of the
Missouri Botanical Garden
e 5. Length of perianth tube of Gladiolus speci
for the species, measured from all known herbarium collections
were captured because their identity is not in question;
deposition were noted because Aeropetes is a legally protected species.
Geometridae: Macaria: Noctuidae: Cucullia, Syngrapha,
potion. Note: the °
been A. convoluli.
'spurge hawkmoth"
reported on G. longicollis by
es (from Goldblatt & Manning, 1998, representing the ranges
) and mouthparts of captured moths. Few butterflies
the few individuals captured were released after sites of pollen
‘amily affiliations: Archiidae: Hypagoptera;
Tychopmoptes; Satyridae: Aeropetes; бй Agrius, Н!
J. M. Wood (in Scott Elliot,
ip-
1891) may also have
Perianth
epidopteran
tube (mm) 1)
Plant species species (r
Position of pollen
Mouthpart
) on insect body
(mm
CLADIOLUS SE CTION от
saunder. Aeropetes tulbaghia
GLADIC OLUS SE CTION | BLANDO S
G. cardinali
'
G.
3
Aeropetes tulbaghia
not measured frons, dorsal thorax
not captured not available
G. е Фе — es tulbaghia not captured not available
G. stefanic -55 — tulbaghia not captured not available
GLADIOLUS — LINEARIFO LIUS
G. guthrie -18 Cucullia inaequalis (1) ca. 16 distal 12 mm of proboscis
G. ile s Hippotion celerio ( 30 distal 25 mm of proboscis
Cucullia inaequalis (1) са. 28 distal 25 mm of proboscis
Cucullia extricata (1) ca. 20 distal 15 mm of proboscis
Macaria simplicilinea (1) <10 no pollen
Hypagoptera sp. <10 no pollen
(
G. nerineoides — les er
GLADIOLUS четох каш
not available
». liliacei JAS Сис us extricata (1) A. í distal 25 mm of proboscis
РА longicollis $ 5-110 Agrius convolvuli (1) 110—115 25 mm from base of proboscis
G. maculatus 23-35 Cucullia terensis (1) 25-30 distal 25 mm of proboscis
Cucullia sp. (1) 5-20 from base of proboscis
G. recurvus 27-36 Hippotion eson (1 10 distal 25 mm of proboscis
z. tristis 10—60 Syngrapha circumflexa (3) 16-17 dorsal thorax and proboscis
Tychomoptes inferior (1) ca. 10 no pollen
lectively making a prominent display, as in G. ner-
ineoides, is well known and has been documented
where (Johnson & Bond, 1994; Goldblatt et al..
1999). The wavelength range of red color exhibited
by Aeropetes flowers is virtually the same as that in
red-flowered species pollinated by sunbirds. There
is no evidence of effective pollination of the but-
terfly-pollinated species of Gladiolus by sunbirds as
there is, for example, in red, tubular-flowered spe-
cies of Tritoniopsis and Watsonia (Iridaceae), Kni-
phofia (Asphodel Erica (Ericaceae),
which may use both sunbirds and Aeropetes equally
as pollinators (Johnson & Bond, 1994; Goldblatt et
al., 1999). i the inclined, and somewhat
flaccid stem and difficulty in reaching the perianth
else
aceae), and
Possibly
tube because of the large, widespread tepals dis-
courage sunbird visits. In connection with Aeropetes
pollination, the record of this butterfly on G. cruen-
tus (Johnson & Bond, 1994) must be mentioned.
From the locality (J. J. Vlok, pers. comm.) this is
clearly for the high-altitude Drakensberg species,
G. flanaganii, and probably reflects the innate at-
traction to red color by Aeropetes. The very fleshy,
stiff cupped tepals, and
of this species (Goldblatt &
suited to sunbird pollination
sturdy flowering stem,
hooded dorsal tepal
Manning, 1998) are
but seem unsuited to butterfly pollination, as does
the nectar, which has a concentration of 35% su-
crose equivalents (Goldblatt et al., 1999). The re-
port of the Nectarinia famosa visiting G. flanaganii
is consistent with the floral adaptations of the spe-
cies, and Aeropetes may simply have been a casual
visitor. Gladiolus cruentus is restricted to lower el-
evations inland from the central KwaZulu-Natal
coast (Goldblatt & Manning, 1998) where Aeropetes
is rare or normally absent, and this red-flowered
species may be pollinated by alternative large but-
terflies, such as Papilio demodocus and Р, nereis,
as has been found for lower-elevation populations
of another red-flowered species of Iridaceae, Hes-
perantha coccinea (Backh. & Harv.) Goldblatt & J.
C. Manning (Goldblatt et al., in press).
DISCUSSION
Although aspects of pollination of Gladiolus spe-
cies by moths and the satyrid butterfly Aeropetes
Goldblatt & Manning 121
Volume 89, Number 1
2002 Gladiolus Lepidoptera Pollination
Noctuid moths photographed while visiting flowers of Gladiolus. А. Cucullia extricata on G. liliaceus.—
g
Fi | 2.
В. Cue "Ша cf.
terensis on G. maculatus (this particular moth was not captured).
closest relatives with different pollination systems
makes it possible for us to draw reasonable conclu-
sions about the evolution of both pollination sys-
tems in the genus, for which a hypothetical phy-
logeny already exists (Goldblatt & Manning, 1998
Moth pollination (Table 6) has evolved in three sec-
have already been documented (Johnson & Bond,
more general terms (Gold-
1994) or discussed 1
[sen Goldblatt et al., 2001).
blatt & Manning,
these two quite different pollination systems are
more fully documented here. Moreover, comparison
of moth- or butterfly-pollinated species with their
=
Analysis of the frequency and taxonomic distribution of night-flying moth and butterfly pollination systems
2 to 5 represent number of species
Table
in E African. species of Gladiolus. Numbers in parentheses in columns
inferred to have the pollination system.
Total Taxonomic tal
Section
pec T sections confirmed
Homoglossum bed (series) and inferred
Pollination
tem Ophiolyza | Blandus Linearifolius Hebea
Moth 0 0 2 (0) 0 (2) 5 (2) 7 З (5) 11 (7%)
Large butterfly 1 (1) З (2) 1 (1) 0 0 5 З (3) 9 (5%)
Annals of the
Missouri Botanical Garden
tions of Gladiolus, Hebea, Homoglossum, and Li-
nearifolius, while butterfly pollination evolved
Gladiolus sects. Blandus, Ophiolyza, and Lineari-
folius. Within section Hebea the two moth-pollinat-
ed species, G. acuminatus and С. robertsoniae, have
such different flowers that we do not believe they
are immediately related. More likely they evolved
from different, bee-pollinated ancestors. Similarly,
the moth-pollinated species of section Homoglos-
sum belong in different series (Tables 1, 2). In ad-
dition, in series Gracilis both floral and leaf mor-
phological differences between G. recurvus, which
has apomorphic leaves with winged margins, like
albens and C.
those of G. gracilis Jacq., and G.
maculatus, which have leathery leaves without
prominent margins or midribs, suggest that G. re-
curvus is not immediately related to the latter spe-
cies pair. Lastly, in Gladiolus sect. Linearifolius, the
only section with both moth- and butterfly-polli-
nated species, species with the two pollination sys-
tems fall in different lineages (Goldblatt & Man-
ning, 1998).
These considerations led to the hypothesis that
moth pollination arose six times within the southern
African species of Gladiolus while butterfly polli-
nation arose at least three times (Table 6). Com-
parison of the pollination systems of immediate an-
cestors of moth- or butterfly-pollinated species
makes it likely that moth pollination arose from
bee-pollinated ancestors in Gladiolus sects. Hebea
and Linearifolius, and in series Gracilis of section
Homoglossum. The sister clade of series Tristis, se-
ries Homoglossum, comprises only bird-pollinated
species, but we cannot say whether the predomi-
nantly moth-pollinated series Tristis evolved from
bird-pollinated ancestors or the putative common
ancestor of the two series, series Teretifolius, which
comprises mostly bee-pollinated species (Goldblatt
& Manning, 1998; Goldblatt et al., 1998a). In trop-
ical Africa, Gladiolus sect. Acidanthera, with seven
species, also shows all of the classic adaptations for
moth pollination. The perianth is white and the
perianth tube 60-150 mm long depending on the
species (Goldblatt, 1996), and flowers produce a
strong odor that, at least for the widely cultivated
G. murielae Kelway (= G. abyssinicus (Brongn. ex
Lemaire) Goldblatt), is strongest at night (unpub-
lished obs.). The flowers of two species of the sec-
tion, б. aequinoctialis Herbert and G. murielae, are
unusual for moth pollination in having dark blotch-
es on the lower tepals, and the entire section Aci-
danthera has an unusual specialization, anthers
with | stiff,
1996). Because of the length of the perianth tube
in species of section Acidanthera, mostly 90-140
acute apical appendages (Goldblatt,
mm long, pollination by sphinx moths (Sphingidae),
which have probosces of comparable length, is in-
ferred but there are no pollination studies for the
Gladiolus species of tropical Africa.
One notable aspect of the evolution of moth pol-
lination in the three southern African sections of
the genus in which it occurs is that these sections
(Gladiolus sects. Hebea, Homoglossum, and Linear-
ifolius) are the only ones in which floral fragrance
is well developed (and is associated primarily with
bee pollination). As we have noted elsewhere
(Goldblatt et al., 19982), floral fragrance is common
in the southern African species of the winter-rain-
fall zone where sections Hebea and Homoglossum
are best represented. It is also present in a few
summer-rainfall zone species of the above sections
that have their closest allies in the winter-rainfall
zone. The scarcity of moth-pollinated species in
summer-rainfall parts of southern Africa may be the
result of the absence of scent in sections of the
genus centered there. In tropical Africa scent in
moth-pollinated species of section Acidanthera ap-
pears to have evolved independently within the sec-
tion (the section as currently defined does not occur
in southern Africa). Floral odor in Gladiolus species
(Goldblatt et al., 1998a) is the result of the com-
bination of many substances and in moth flowers
perceived odor represents different mixes and pro-
portions of these same series of compounds.
Apart from fragrance, moth flowers differ from
their bee-pollinated ancestors in perianth tube
length, color, and size of the tepals. Large size is
probably important for display purposes at night as
is the pale coloration. An unusual aspect of moth-
pollinated flowers in Gladiolus is the frequency of
dull patterning or mottling on a pale background,
most notable in G. hyalinus, G. liliaceus, G. ma-
culatus, and even some forms of G. longicollis (all
sect. Homoglossum) and in G. emiliae and G. guth-
riel (sect. Linearifolius). Dark mottling is less pro-
nounced in G. hyalinus and G. liliaceus at night, as
the pigments either change in intensity or color or
both. The mottling may be camouflage, protecting
flowers from predation during the day, as Johnson
(1994) has suggested for the maroon-flowered or-
chid, Monadenia ophyridea, which is presumably
located by its settling moth pollinators by scent
alone.
Butterfly-pollinated flowers in Gladiolus sect.
Blandus appear have evolved from ancestors
adapted for pollination by long-proboscid flies, the
predominant pollination system in the section, and
the one present in species most closely allied to the
clade of butterfly-pollinated species in the section
Goldblatt & Manning, 1998, 1999). Long-probos-
—
Volume 89, Number 1
2002
Goldblatt & Mannin
g 123
Gladiolus Lepidoptera Pollination
cid fly flowers have similarly long floral tubes and
differ
cream or pink with red nectar guides often outlin-
most significantly in floral pigmentation,
ing a white zone. In Gladiolus sect. Ophiolyza the
ancestors of the two butterfly-pollinated species
have flowers adapted for pollination by sunbirds
1999). In
series Linearifolius butterfly-pollinated б. nerineo-
(the С. dalenii complex) (Goldblatt et al..
ides and G. stokoei (putatively butterfly pollinated)
fall within a clade with the bee-pollinated G. brev-
ifolius and long-proboscid fly pollinated G. monti-
cola (Goldblatt & Manning. 1998,
At first it might appear that a simple shift from
ancestral pink or cream flowers to red ones may
effect a shift to butterfly pollination. In Gladiolus
increase in flower size appears equally important.
Moreover, other more subtle changes suggest that
more complex genetic adaptations are necessary.
including a shift in the position of the anthers and
stigmatic surfaces, either prominently exserted or
included in the floral tube. and a shift in flowering
time (butterflies are on the wing only in late sum-
mer).
vegelative and physiological adaptations would also
be required to produce flowers in an area of sum-
mer heat, drought, and low atmospheric humidity.
a time that necessitates the separation of vegetative
growth from the flowering phase of the life cycle.
Johnson and Bond (1994) have pointed out a fea-
ture of butterfly pollination in the southern African
that
question. This is that so many butterfly-pollinated
winter-rainfall zone relates. directly to this
species of the area are rare, or are highly local
endemics. We suspect this has nothing to do with
the intrinsic effects of having a specialist pollina-
tion system and is correlated instead with the fact
that species flowering in the late summer and au-
tumn in an area of mostly winter precipitation re-
quire highly specialized habitats that are by their
Thus. б.
vives summer drought by growing along watercours-
very nature highly local. cardinalis sur-
es in cool mountain habitats. often in the spray of
perennial waterfalls. Gladiolus sempervirens, G. ste-
faniae, and G. stokoei also grow in montane habitats
and are restricted to slopes with poor drainage
where the summer southeast winds bring moisture
in the form of fog that keeps the ground moist al
this otherwise dry season. Gladiolus nerineoides is
restricted to shaded south-facing cliffs where mois-
ture percolating from the slopes above keeps their
substrate relatively damp. Even in these special-
ized habitats vegetative specialization makes sur-
vival for these summer-flowering species possible.
Gladiolus nerineoides, G. stefaniae, and G. stokoet
produce foliage leaves in the winter, after the wet
In the winter-rainfall zone a whole series of
season has begun. Their flowering stems bear short,
often entirely sheathing, and sometimes dry leaves
while the foliage leaves are dry and often lost by
the time flowers are produced.
An aspect of evolution of specialized pollination
systems is a change in the nature of the reward.
Bee-pollinated species of Gladiolus mostly produce
sucrose-rich or sucrose-dominant nectar of relative-
1998a). The
in moth-pollinated flowers is similar, al-
This
pattern is one that is common in the Iridaceae. In
(Goldblatt
et al., in press) nectar sugars usually have concen-
ly high concentration (Goldblatt et al.,
pattern
though sugar concentration is often higher.
moth-pollinated species of Hesperantha (€
trations of 45-50% sucrose equivalents. The rela-
tively high volume of nectar in most moth flowers,
especially compared to their presumed bee-polli-
nated ancestors, may relate to the higher nutrient
needs of active, fast moving moths with relatively
arge bodies, especially in the case of Sphingidae.
which hover while feeding.
Unlike nectar of moth flowers, the nectar of but-
terfly-pollinated Gladiolus species shows a lower
sugar concentration than their presumed long-pro-
boscid fly or bee-pollinated ancestors (Goldblatt et
al.. 1998a: Goldblatt & Manning, 1999), but a rel-
atively high volume. Exactly why the lower nectar
sugar concentration is characteristic of large but-
terfly pollination is uncertain. The higher viscosity
of concentrated nectar may make it difficult for an
insect to draw through the proboscis, but this is
evidently not a consideration for small arge
moths. which have similarly constructed and some-
times equally long and narrow mouthparts. Another
feature of butterfly nectars in Gladiolus is the shift
to higher proportions of hexose sugars. While most
species of section Blandus have the sucrose-dom-
inant type nectar found in related long-proboscid
fly pollinated species (Goldblatt & Manning, 1999),
G. insolens has hexose-rich nectar. A similar but
less pronounced shift is evident in the two butterfly-
pollinated species of section Ophiolyza, which have
sucrose-rich nectar with nearly equal proportions
of sucrose and hexose sugars, and in G. nerineoides.
which shows the same shift, while related species
=
have nectar with higher proportions of
(Goldblatt € Manning, 1999; Goldblatt et al..
1999). Such changes in nectar sugar chemistry are
sucrose
rare in the Iridaceae, especially in the subfamily
Crocoideae (syn. Ixioideae) to which Gladiolus be-
longs. and are presumably pollinator driven, re-
flecting a preference of Aeropetes for nectar with
elevated levels of hexose sugars. rather than a ran-
dom change.
Shifts in pollinators are common within the larg-
124
Annals of the
Missouri Botanical Garden
er genera of the African Iridaceae and are an im-
portant factor in the radiation of the family. Else-
where we have speculated that as many as 32 shifts
in pollination system have occurred just in Gladi-
olus in southern Africa (Goldblatt et al., 2001). This
represents an average of one shift for every five
species. The convergent evolution of moth polli-
nation at least six times and butterfly pollination
three times reflects the evolutionarily labile floral
and vegetative morphology of this highly successful
genus. As outlined above, shifts that may appear
quite simple, appearing to represent changes main-
ly in pigmentation and floral tube length are, in
fact, complex and represent a closely integrated se-
ries of adaptations that include, in addition to floral
patterning, shifts in fragrance characteristics, tim-
ing of the opening and closing of the tepals and
sometimes scent release, nectar physiology, season-
al phenology, and associated vegetative adaptations
for those species of the southern African winter-
rainfall zone that flower in summer. The develop-
mental complexity associated with these shifts
makes their high frequency seem even more im-
pressive.
Literature Cited
Baker, H. G. . Baker. 1983. Floral nectar me
stituents in re — to pollinator type. Pp. 117—141 in
C. E. Jones & R. J. Little (editors), Такі of Ex-
perime а) seme ate Biology. Scientific and Academic
Editions, New
Bue mad S.L. 1983. Buzz pollination in е rms.
73-113 in C. E. Jones & R. J. Li (editors).
Handbook of Experimental. Pollination. E Nostrand
Reinhold, New York
Faegri, K. & L. van der Pijl. 1979. The Principles of
ation Ke sls 3rd ed. Pergamon Press, New Yo
Goldblatt, Р. 19 "ail in Tropical Africa. Timber
Press, pora. Oe on.
& J. C. Man 1998. Gladiolus in Southern Af-
rica. Fernwood Press, Town.
——.. 1999. The long-proboscid fly pollina-
. Ann. Missouri Bot.
æ
=
lion system in E (Iridaceae).
Gard. 86: 758-774
Volume 89, Number
‚ 2000. The long-proboscid fly pollina-
tion system in southern Africa. Ann. Missouri Bot.
Gard. 87: 146-170.
Bari & J. C. Manning. 1991. Sulcus variability
in че pollen grains of pow ‘eae subfamily Ixioideae.
Ann. Missouri pot — 950-961.
& | — 1995. Pollination in Lap-
etrousia subgenus — (Iridac eae: Ixioideae).
in. Missouri Bot. Gard. 82: 517-5:
› Bernhardt & J. C. Manning. 1998a. Adaptive
аса of bee pollinated Gladiolus species (Iridaceae)
in southern Africa. Ann. Missouri Bot. Gard, 85; 492
517.
E
— ——. 1998b. Pollination by mon-
key Бева (5с — Hopliini) in peta taloid geo-
phytes i in southern Africa. Ann, Missouri Bot. Gard. 85:
215-230.
1999, Evidence of bird pol-
lination in "s iva 'eae of southern Africa. Adansonia
er. oO, . 2:
‚ 2001. Radiation of pollina-
tion — in КҮ (Iridaceae: Crocoideae) in
southern — Ann. Missouri Bot. Gard. 88: 713—734
„Р. Bernha rdt & J. C. Manning. In press.
Floral biology P сони (Iridaceae: Crocoideae):
How minor spud 1 floral presentation change the pol-
lination pA Ann. Missouri Bot. Gard.
Johnson, S. D. Evidence for Batesian mimicry in a
butterfly- e orchid. Biol. J. Linn. Soc. 53: 91—
104
T
—
©
. Moth pollination of the cryptic — or-
chid, Monadenia ophrydea. Flora 190: 105-10
—— & V. J. Bond. 1994. Red flowers and Iter
pollination i in the fynbos of South Africa. Pp. 137-148
. Arianoutsou & R. Grooves (editors), do Animal
ner ‘tions in oe Type Ecosystems. Kluwer
— — Jordrecht.
"M user, R. ne Scent of Orchids. Elsevier, Amster-
dar
dan
— J. C., P. Goldblatt & P. D. J. Winter. 1999, Two
new species of Gladiolus (lridaceae: Ixioideae) from
South Africa and notes on pi probosc ‘id fly pollination
in the genus. Bothalia 29: —223.
Ogden, E. C., G. S. Raynor, J. x Hayers & D. M. Lewis.
1974, e of Sampling Airborne Pollen. Hafner
2
Pre
Scott — "1891. Notes on the fertilisation e South
Afric 'an — E an flowering plants. Ann. Bot. 5
05
l, pp. 1-124 of the ANNALS OF THE MISSOURI BOTANICAL GARDEN
was published on April 15,
2002.
а ра
ANNALS OF THE MISSOURI BOTANICAL GARDEN
AND MISSOURI BOTANICAL GARDEN ANNUAL REPORT
ARE NOW AVAILABLE IN JSTOR!
JSTOR’, a not-for-profit organization, is an important endeavor
dedicated to helping the scholarly community take advantage of advances
in electronic technologies. The JSTOR Archive makes available the
complete backfiles of Annals of The Missouri Botanical Garden,
Missouri Botanical Garden Annual Report (1890-1912), and more than 100
other scholarly journals to researchers through participating libraries.
JSTOR has recently announced the release of its Ecology & Botany
Collection. Developed in cooperation with the Ecological Society of America,
the Collection includes the archives of twenty-nine journals, offering
online access to more than 1 million pages of important historic literature.
The Ecology & Botany Soper
American Journal of Botany : International Journal of Plant Sciences |
American Midland Naturalist | Journal of Animal Ecology e
American Naturalist | ` . Journal of Applied Ecology
Annals of the Missouri Botanical — Journal of Biogeography
Annual Review of Ecology and Systematics — Journal of Ecology ee
Biotropica Journal of Tropical Ecology
Brittonia Journal of The Torrey Botanical —
Conservation Biology Limnology and Oceanography
Diversity and Distributions Missouri Botanical Garden Jab дерек
Ecological Applications _ New Phytologist -
Ecological Monographs - : . Paleobiology . —
Ecology ets as . Quarterly Bien iof Biology.
Evolution : ; Systematic Biology '
Functional Ecology Үе tur улты puer
Global Ecology and 1 Biogeography 1 Letters m
{ —— fe PE Ж, 1 a Ж
ISTOR 1s available at nup./
г Y А
ишо A oJ
o 3 : _ F J
120 Fifth Avenue, New York, NY 10011.
— — —
CONTENTS
A Systematic Revision of Breonia (Rubiaceae—Naucleeae)
Sylvain G. Razafimandimbison
Classification, Origin, and Diversification of the New Zealand Hebes (Scrophulariaceae)
Steven J. Wagstaff, Michael J. Bayly, Philip J. Garnock-Jones & Dirk С. Albach
Phylogenetic Reconstruction of the Neotropical Family Quiinaceae (Malpighiales)
Based on Morphology with Remarks on the Evolution of an Androdioecious Sex
Distribution Julio V. Schneider, Ulf Swenson & Georg Zizka
A Review of the Genera Roentgenia and Potamoganos (Bignoniaceae) — Warren D. Hauk
Molecular Data Indicate Complex Intra- and Intercontinental Differentiation of American
Draba (Brassicaceae) Marcus Koch & Ihsan A. Al-Shehbaz
Evidence for MN and Butterfly Pollination i in Gladiolus (lridaceae—Crocoideae) .....
Peter Goldblatt & John C. Manning
Annals '
of the
Missouri
Dotanical
"mo"
umber
Volume 89, Number 2
Spring 2002
Annals of the
Missouri Botanical Garden
The Annals, published quarterly, contains papers, primarily in systematic botany,
contributed from the Missouri Botanical Garden, St. Louis. Papers originating out-
side the Garden will also be accepted. All manuscripts are peer-reviewed by quali-
fied, independent reviewers. Authors should write the Managing Editor for informa-
tion concerning arrangements for publishing in the ANNALS. Instructions to Authors
are printed in the back of the last issue of each volume and are also available online
at www.mobot.org/mbgpress.
Editorial Committee
Victoria C. Hollowell
Editor, |
Missouri Botanical Garden
Amy Scheuler McPherson
Managing Editor,
Missouri Botanical Garden
Diana Gunter
Associate Editor,
Missouri Botanical Garden
Aida Kadunic
Senior Secretary
Barbara Mack
Administrative Assistant
Ihsan A. Al-Shehbaz
Missouri Botanical Garden
Gerrit Davidse
Missouri Botanical Garden
Roy E. Gereau
Missouri Botanical Garden
Peter Goldblatt
Missouri Botanical Garden
Gordon McPherson
Missouri Botanical Garden
P. Mick Richardson
Missouri Botanical Garden
Charlotte Taylor
Missouri Botanical Garden
Henk van der Werff
Missouri Botanical Garden
For subscription information contact ANNALS
OF THE MISSOURI GARDEN, % Allen Marketing
& Management, Р.О. Box 1897, Lawrence, KS
66044-8897. Subscription price for 2002 is
$140 per volume U.S., $150 Canada & Mexico,
$175 all other countries. Four issues per vol-
ume. The journal Novon is included in the sub-
scription price of the ANNALS.
annals@mobot.org (editorial queries)
http://www.mobot.org/mbgpress
© Missouri Botanical Garden 2002
+h
THE ANNALS OF THE MISSOURI BOTANICAL
GARDEN (ISSN 0026-6493) is published quar-
terly by the Missouri Botanical Garden, 2345
Tower Grove Avenue, St. Louis, MO 63110.
Periodicals postage paid at St. Louis, MO and
additional mailing offices. POSTMASTER: Send
address changes to ANNALS OF THE MISSOURI
BOTANICAL GARDEN, % Allen Marketing &
Management, Р.О. Box 1897, Lawrence, KS —
66044-8897,
The mission of the Missouri Botanical Gadi is to discover end share knowledge about plants and
their e "nt, in order to preserve and enrich life.
7 us | @ mis paper meets the requirements of ANSI/NISO — (Permanence of Paper).
olume 89
umber 2
002
nnals
of the
ISSOUTI
otanica
arden
INTRODUCTION:
CONSERVATION AND THE
FUTURE OF LIFE!
George E. Schatz? and
P. Mick Richardson?
Upon initial thought, conservation might seem an
anomalous toplc for a systematics symposium,
However, although systematics is often perceived
only in its arcane, classificatory function, and con-
f
dwindling populations of charismatic large animals,
servation has primarily had a publie persona o
they are in fact two sides of the same coin, Or said
another way, they concern themselves with a com-
the diversity of life on Earth.
Annual Sys-
tematics Symposium was to highlight the intimate
mon Currency, I...
Indeed. one of the goals of the 47th
connection between systematics and conservation,
Vs scientific disciplines, both encompass broad ap-
proaches, from the “top-down” landscape and
ecoregion paradigms of conservation, and the “Tree
of Life” phylogenetic perspective in systematics, to
“bottom-up” Population Viability Analyses and ex
situ interventions in conservation, and analyses of
genetic variability in systematics. At all levels, sys-
tematies directly informs conservation. and all per-
spectives ultimately converge upon species, the
most recognizable unit of biodiversity.
Among the countless species in the cornucopia
of biodiversity, one species—our own—has distin-
guished itself from all others. Human beings have
the dubious distinction of having become the most
success. we have «on: the first species to alter
ı achieving that
—
successful species of all time. hi
basic biosphere processes since the original pho-
l
sphere and sent evolution down an aerobic path.
tosynthetic bacteria added oxygen to the atmo-
Although the rise of one species has undoubtedly
often contributed to the decline of another species
during that long evolutionary continuum, our suc-
cess now directly or indirectly negatively impacts
(or eventually will impact) nearly all of the other
10-100 million species with whom we share our
home. We are both the problem. and the only po-
tential solution.
But as К. O. Wilson (2002) so eloquently artic-
ulates in his most recent book, The Future of Life.
the
The following Sympo-
there is much about which to be hopeful in
urgent search for solutions.
sium contributions all embrace that optimism. il-
» "bottom-
Joel
syste matics
lustrating the full range of “top-down” ti
ultimately human-centered approaches.
the
during the past several decades and its ensuing di-
up.”
Crac тай invokes re naissance e in
! This
Missouri Botanic ‘cal Garden. The symposium was held i
> National Science Foundation unc
»eaker
ect a fine 5
Diana Gunter for
? Missouri Botanical Garden. P.O.
their editing and political skills, and th
Box 299, St. Louis.
ANN. MISSOURI Bor.
and the six articles that follow it constitute ys i
Kathy Hurlbert anc d
|" Myers for his ett used on the cover of the symposium brochure, Victoria Hollowell. Amy
ceedings of и | nh no ıl Systematics Symposium of the
ouis “Miss t October 2000. The symposium was
grant DEB- ‘998 1642 2. PMR thanks Peter Raven for helping
xpert staff for organization and уто sate skills.
IcPherson. and
St. |
ler
—
e spec akers for bei "ng suc E an inte resting group of scientists
Missouri 63166-0299, U.S.A.
GARD. 89: 125-126. 2002.
126
Annals of the
Missouri Botanical Garden
rect application to societal problems as evidence of
its relevance in conserving and sustainably using
biodiversity. Both he and George Schatz emphasize
the fundamental importance of the natural history
collections that underpin all our knowledge of bio-
diversity, and underline their critical role in di-
rectly informing conservation planning using pow-
erful new informatics tools. The mapping of
biodiversity shows us that most species are not ran-
domly distributed over the face of the earth, and
that collectively they form distinct associations, of-
ten highly restricted in area, that correlate with the
variety of life-supporting milieu and historical phe-
nomena of isolation and dispersal. Thus, David Ol-
son and Eric Dinerstein advocate the landscape
paradigm to ensure functional representatives of
each of the world’s distinct ecoregions, i.e., those
that satisfy the minimum size requirements of all
their constituent species, and retain the potential
for migratory processes.
In virtually all of those ecoregions, we are one
of the constituent species, often the dominant опе;
for better or, now, mostly worse, all of the rest of
the species are consigned to live alongside us.
Again, we are the only potential solution. Gary
Nabhan and his colleagues paint a portrait of the
Colorado Plateau in which indigenous cultural di-
versity and biological diversity reciprocally sustain
one another, serving as an example for the local,
community-based management of biodiversity. De-
bra Moskovits and her colleagues present an in-
spiring vision of Chicago Wilderness that serves to
resurrect our innate capacity for stewardship amidst
even our most dense, urban environments. In con-
trast, Stuart Pimm validates the alternative scenar-
io, dispelling once and for all any lingering myths
that we are not the current agents of extinction,
which indeed we regrettably are.
The consequences of our success, and the ulti-
mately self-detrimental manner in which we cur-
rently live, are becoming increasingly clear. As we
strive to disseminate our emerging knowledge, we
must nurture the belief that it will soon foster the
ethical principles and moral standards by which we
will fully embrace not only each other, but all of
our fel low species.
Literature. Cited
Wilson, E. O. 2002. The Future of Life. Alfred A. Knopf,
New York.
THE SEVEN GREAT Joel Cracraft?
QUESTIONS OF SYSTEMATIC
BIOLOGY: AN ESSENTIAL
FOUNDATION FOR
CONSERVATION AND THE
SUSTAINABLE USE OF
BIODIVERSITY"
ABSTRACT
The three missions of Systematics Agenda 2000 (SA2000)—to inventory Earth's species, to understand their rela-
tionships, and to use the latter to create predictive information systet ns—define an agenda of research for systematic
biology. The recognition that systematic knowledge underpins biologie ‘al knowledge in general, and applied biology i in
a | 8 | е
particular, has resulted in an amazing growth in systematics over the past decade. Increasingly, systematics is being
used to solve societal problems. This paper describes seven great questions within systematics and — 's thelr ‘ir
relevance for, and contribution to, conserving and sustainably using biodive sily. These questions fall into four broad
categories: Questions about diversity: What is a species? and How many species are there? Questions about phylogeny:
What is he Tree of Life? and What has been the history of character transforma in? Questions about biogeograph y:
| geograph
Where are Earth's species distributed? and How have species! distributions changed over time? and Questions about
I 8
phyloinformatics: How is phylogenetic history predictive?
Key words; biogeography. diversity, inventory, phylogenetics. phyloinformatics, systematics, Systematics Agenda
2000
Systematic biology occupies the central core of tion of the importance of systematics has never
biodiversity science. The four great themes of sys- been stronger.
tematics—diversity, phylogeny, biogeography, and There is a sense, however, in which some of
classification (to be subsumed here into a broader these discussions linking systematics and conser-
context termed. phyloinformatics)—form a critical vation can be said to be too narrow, both from the
—
foundation for all other disciplines of biology. The perspective of syste matic s and from conservation.
discovery and naming of life's diversity. its evolu- Thus, "conservation," a
east in the expansive
tionary history, its distribution across Earth to form sense it will be used here, is more than just efforts
habitats and ecosystems, as well as how compara- to save endangered species and ecosystems. or to
tive information about organisms can be organized create protected areas. While certainly embracing
in a predictive manner, underpin, to one degree or these objectives, conservation in the real world cov-
another, all biological knowledge. ers much more ground philosophically, as well as
Beginning in the early 1990s systematists be- in practice. In a world losing its biodiversity at an
came increasingly engaged in conservation and the accelerating rate, systematics needs to be seen as
role played by systematics in solving societal prob- a crucially relevant and important science in meet-
lems (Systematics Agenda 2000, 1994a, 1994). ing the challenges of global environmental change
Since then, systematists have published numerous (sensu lato), at the same time contributing strongly
papers that address the contributions of systematics to traditional conservation practice.
to conservation biology, focusing especially on such It will therefore be a theme of this paper that
topics as diversity patterns, species concepts, geo- systematics and systematists should be approaching
graphie distribution, and endemism. The recogni- conservation as a subset of solutions that must be
! I thank Peter Raven and P. Mick Ric ен for asking me to participate in the 47th Annual Systematics 5
at the Missouri Botanical Garden. Meredith Lane ea many helpful pea on the manuscript that greatly —
its clarity; I am most grateful to her. Many | of the intellectual seeds of this article derive from the hard work and
brilliance of the höndreda of systematists who ud to Systematics Age — 2000. which started in the United
States but has now spread у vorldwide. The initial effort of Systematics Agenda 2000 was supported by a grant from the
U. S. National Science Foundation (DEB-9396035). Consider this yet another contribution from that investment.
? Department of Ornithology, American Museum of Natural History. New York, New York 10024, U.S.A. jle@amnh.org
ANN. Missouni Bor. GARD. 89: 127-144. 2002.
128
Annals of the
Missouri Botanical Garden
implemented to mitigate the adverse effects of en-
vironmental impacts in general, all the time real-
izing that the contributions undertaken to sustain
and improve human well-being can themselves be
seen as a conservation initiative. Stemming the loss
of biodiversity is essential, but the factors leading
to that loss are imbedded in a complex causal nex-
us that encompasses all aspects of society. People
across the globe use tens of thousands of species
to sustain their lives in one way or the other, and
the causal complexity of that use is what makes
saving biodiversity so difficult—it cannot be sepa-
rated from how societies function. This reality is
encapsulated in the activities of the Convention on
Biological Diversity (CBD) and many nongovern-
mental organizations. Effective conservation efforts
cannot be separated from the elimination of poverty,
which cannot be separated from the improvement
of women's health, education, and economic en-
franchisement, which cannot be disentangled from
governmental policies of many kinds, and on and
on. New
necessary to meet the many environmental chal-
nowledge about the world (science) is
lenges created by human activities, and thus sys-
tematics can help in all sorts of ways not generally
realized by most practicing systematists or other
biodiversity scientists. This is what I mean by the
need for systematists to have a more expansive vi-
sion for their discipline and for its importance.
This paper explores the contributions of system-
atics to science and soc let y, firstly, by summarizing
some of the literature published since the mid-
1990s, when participants in Systematics Agenda
2000 (842000) released many of their assessments,
and secondly, by pointing to new developments that
will become increasingly important in the future.
This will be done within the framework of what
might be called the seven great questions of sys-
tematic biology. which to me at least, encapsulate
most of the research agenda, and importance, of
systematics to society. Not all of these questions
will be treated equally here. inasmuch as some
have been the subject of a substantial recent lit-
erature (see citations below).
SYSTEMATICS AGENDA 2000: BRINGING
SYSTEMATICS TO SOCIETY
In the early 1990s the Society of Systematic Bi-
ologists, the American Society of Plant Taxono-
mists, and the Willi Hennig Society, in cooperation
with the Association of Systematics Collections,
launched an effort to document the importance of
systematics to society (Anonymous, 1991). Through
its many publications, Systematics Agenda 2000
transformed the way systematists view their own
discipline, and it helped create an atmosphere in
many countries throughout the world in which sys-
tematics gained in stature and importance as ап
integral component of biodiversity science.
Jecause many systematists, both young and old,
might be unfamiliar with the rich panoply of pub-
lished results of this effort, these are summarized
1 Table 1. These papers cover a broad range of
subjects that articulate the importance of system-
atics. Collectively, they have reached out to system-
atists and biodiversity science policymakers around
the world and have been responsible, in varying
degrees, to promoting and building systematics.
Thus, to take one example, following SA2000 sym-
posia at the Royal Society in London (Claridge,
1995) and the French Academy of Sciences in Par-
is, new initiatives were formed to promote the ob-
jectives of systematics across Europe (Blackmore &
Cutler, 1996
Systematics Agenda 2000, although begun in
North America, was developed as a global effort.
Thus, the core documents of SA2000 were inten-
tionally not copyrighted so that they could be taken,
and used, by systematists to promote systematics
research and capacity building anywhere. Transla-
tions were encouraged, and that strategy worked as
numerous individuals and groups adopted the lan-
guage and content of SA2000 in their efforts (Table
). Today, SA2000 is truly international. System-
atics Agenda 2000 International (SA2000I) is a
program of the International Union of Biological
Sciences (IUBS) and functions as the systematics
core element of the international biodiversity sci-
ence program DIVERSITAS | [DIVERSITAS|.
Through DIVERSITAS, SA20001 has been active
in providing advice on systematics science and ca-
pacity building to the Convention on Biological Di-
versity (CBD) and its Global Taxonomy Initiative
[GTI]. Some of the SA2000I/DIVERSITAS docu-
ments in support of the GTI are available on the
CBD website; these documents contain recommen-
dations that have been adopted by the CBD Con-
ference of the Parties (the 181 countries that have
ratified the Convention). (For the record, as of 12
September 2001, only six signatory countries to the
CBD have failed to ratify: Afghanistan, Yugoslavia,
Thailand, Tuvalu, Kuwait, and the United States of
America.)
THE SEVEN GREAT QUESTIONS OF SYSTEMATICS
The scientific agenda of systematics and its rel-
evance for society occupy four great themes: diver-
sity, phylogenetic history, biogeography, and clas-
Volume 89, Number 2 Cracraft 129
2002 Systematic Biology
An overview of the papers produced by participants in Systematics Agenda 2000 over the last decade.
Table
‘uss aspects of building systematics
These publications treat the role and importance of systematics to society, disc
science capacity, or address policy issues involving systematics and biodiversity.
Publication Theme
General SA2000 Documents
Anonymous (1991) Announced formation of SA2000
Systematics Agenda 2000 (1994a) Color brochure disc — — e of systematics
and describing SA2000's three missions
Systematics Agenda 2000 (1994p) Technical report hw ail for 1994a
BioScience 1995 (vol. 45, no. 10)
Simpson & Cracraft (1995)
Savage (1995)
Miller & ер (1995) Role of systematics in agriculture
Brooks et al. )5) Role of systematies in ecology and behavior
Lauder et al. Lo
Overview of systematics and papers in BioScience
Role of systematics in biodiversity science
Role of systematics in comparative morphology and
physiology
Davis (1995) Systematics and public health
Biodiversity pie Conservation 1995 (vol. 4, no. 5)
Claridge (1995 Introduction to SA2000 symposium held at the Royal
Society, 12 April 1994
Overview and history of SA2000
Eshbaugh (1995)
Building systematics and biodiversity science capacity
Cracraft (1995)
Wheeler (19054)
Prance (1995
Biodiversity and systematics inventorie
Systematics, conservation, and DAS develop-
теп!
Jones (1995) BioNet-International and capacity building
McNeely (1995)
Annals of the Missouri Botanical Garden 1996
(vol. 83, no. 1)
Richardson (1996)
Systematics and conservation
Introduction to SA2000 Symposium, 41st Annual Sys-
tematics Symposium, MOBOT
Monson (1996) Phylogenetics and comparative plant physiology and
development
Rossman & Miller (1996) Role of systematics in agriculture and forestry
Vecchione & Collette (1996) Role of systematics in fisheries and marine
biodiversity
Oliver (1996) Importance of systematics for public health
ane-Wright (1996) Role of systematics in conservation
Balick (1996) Systematics and ethnobotany
Other publications
Sy — Age nda 2000 (1994)
Wheeler (19
French translation of SA2000 (1994a)
pe and biodiversity policy
Agenda — 2000 (1996) n translation and expansion of SA2000 (1994)
E. — & Cutler (1996) dr SA2000 in —
r & Cracraft (1996) Building systematics capac
Workshop report, cosponsore sd by SA2000 Internation-
. on Convention of Biological Diversity Global
lI
IL
—
Museum of Natural History (1999 ا
р;
Taxonomy Initiativ
Baas (1999) SA20001, DIVERSIT AS, and the CBD
Cracraft (2000) Building systematics capacity
sification. Each of these four themes leads to a THE FIRST GREAT QUESTION: WHAT IS A SPECIES?
series of fundamental questions—the seven great
questions of systematics. These questions broadly
cover what systematists do, and are why, in my
opinion, systematics is the central, undeniable core
No question, probably, has generated more con-
troversy, been so opaque to solution, and yet re-
mains as crucial and important today as it ever has.
of biodiversity science. than “What is a species?" In systematics, which is
130
Annals of the
Missouri Botanical Garden
a taxon-based science, it is unquestionably the
most fundamental question that can be asked. As-
tonishingly, many biologists profess to be tired of
the controversies over this question, yet instantly
remain willing to engage in the debate, whether or
not they themselves are professionally concerned
with individuating species limits, i.e., whether they
are systematists or not. Everyone, it seems, knows
what a species is, or more likely, what is not!
The reason for the dispute is fairly obvious: at
some level, notions of species are at the very heart
of biological understanding and so there is a lot at
stake. Species concepts intersect with a whole se-
ries of conceptual and disciplinary agendas, from
those of systematists who discover and describe
Earth's diversity, to those reconstructing phylogeny
and biogeography, to those attempting to under-
stand the origins of species from a genetic or eco-
logical viewpoint, to those interested in conserva-
tion, ecology, and applied biology. Species are the
basic units of systematics, evolutionary biology, and
diversity. Thus, species concepts touch upon ele-
mental philosophical arguments about the reality of
the units of nature as perceived by biologists of
varying disciplines as well as scientific. opinion
about how life evolves.
Given this crucial importance, it is ironic that
there is so much divisiveness over “What is a spe-
cies?” You would think biologists could have solved
this issue by now. Yet each year brings forth nu-
merous papers on the subject, and books keep flow-
ing (most recently, for example, Ereshefsky, 1992;
Kimbel & Martin, 1993; Claridge et al., 1997;
Howard & Berlocher, 1998; Wilson, 1999; Wheeler
& Meier, 2000), all with no diminution of differ-
ences of opinion.
Nevertheless, it can be argued that over the last
decade, practicing systematists—those biodiversity
scientists whose research most directly bears on
this issue—are gradually converging upon a com-
mon solution to this question, either as a result of
` both.
These systematists see species as basal elici
practical or theoretical considerations,
clusters of diagnosably distinct populations (groups
of individual organisms). Most of these systematists
do not endorse a particular concept of species; they
go about their work inventorying, describing, and
monographing without a heavy burden of theorizing.
But if one had to put a name on their concept, it
would appear to be most similar to that advocated
by supporters of a phylogenetic species concept
(Nelson & Platnick, 1981; Cracraft, 1983, 1989a;
Nixon & Wheeler, 1990; Wheeler & Platnick,
2000
The reason why this is the most common ap-
proach to species is very practical, as well as sim-
ply a matter of numbers: the majority of systema-
tists working on invertebrates, especially insects,
many plant groups, and various vertebrate taxa
view species as basal diagnosable units because
this best fits the way they partition the taxonomic
variation they have observed. How could this not
be the case when the large majority of described
species are known from a handful of specimens,
often single individuals? Or when nothing is known
about their biology or patterns of genetic variation?
Thus, the debate over species, in a practical
sense, comes down to a choice between a phylo-
genetic species concept (PSC) and a biological spe-
cies concept (BSC). The latter, surprisingly, is ap-
plied by very few practicing systematists who
inventory and describe species, but has strong sup-
port from advocates of the evolutionary systematics
of Mayr (1942, 1963, 1982, 2000) and by those
whose interests lie with population genetics (Coyne
1988; Avise & Ball, 1990) or evolutionary
biology in general (e.g., Bock, 1987; Futuyma,
998). The interchanges among advocates of the
PSC and BSC have been incessant. It is not my
purpose to review these debates as the central ar-
guments and positions can be found in Wheeler
and Meier (2000). Instead, pursuant to the theme
of this paper, | comment briefly on the relevance
et al.,
of species concepts to conservation and applied bi-
ology.
Species concepts are important because they al-
low us to propose hypotheses about the ontology of
nature: different species concepts generally imply
a different ontology (Cracraft, 1987, 1989b). This
is not just philosophical mumbo-jumbo. One needs
a clear idea of the entities of nature so that one can
count and describe patterns of diversity, as well as
understand how entities behave (i.e., how they par-
ticipate in processes). Thus, there are very practical
consequences stemming from the adoption of a par-
ticular species concept, especially in the descrip-
tion and enumeration of diversity.
Conservation biology, although inherently cross-
disciplinary, emerged primarily from an ecological
tradition, and ecologists and other non-systematist
biodiversity scientists have come to their under-
standing of species and speciation through their
formal university training in ecology, genetics, and
evolution. Most of that training has accepted the
BSC because it has been the canonical view of spe-
cies since the early influential work of Ernst Mayr
1942, 1963) and because it has been followed in
most contemporary textbooks (e.g., Futuyma, 1998).
However, many conservation biologists began to see
a problem with applying the BSC because of its
—
Volume 89, Number 2
2002
Cracraft 131
Systematic Biology
ambiguous treatment of discrete taxonomic varia-
tion. Thus, under the BSC, diagnosable populations
might be ranked either as a species or subspecies,
or subspecific rank itself might be applied to diag-
nosably distinct forms as well as to arbitrary sub-
divisions of clinal variation,
In an influential paper, Ryder (1986: 9710) sum-
marized the opinions of conservation biologists
working within the zoo community: “Out of a sense
of frustration with the limitations of current mam-
malian taxonomy [broadly using the BSC] in deter-
mining which named subspecies actually represent
significant adaptive variation, those assembled at
the Philadelphia conference [of zoo biologists | will-
ingly discarded the concept that all subspecies are
equal. Rather, it emerged that zoos ought properly
to address the conservation of evolutionary signifi-
cant units (ESUs within species).” They went fur-
ther to suggest that identification of ESUs be made
on the basis of concordance across multiple kinds
of data and that “when geographic distribution data
indicate the existence of discrete [italics added |
populations within the range of a species. an esti-
male of genetic distance, for example, should be
made to determine whether the populations have
ESL
valionists were getting at: the traditional BSC ap-
status.” It is fairly clear what these conser-
proach to individuating units of nature was not
working for their purposes. While the determination
of whether a population could be judged to have
"significant adaptive variation" verges on nonoper-
ationalism. the key recommendation of their pro-
identifying discrete populations—does not.
posal
They grasped the reasonable idea of trying to con-
serve and manage (in situ and ex situ) diagnosable
and distinct populations, ESUs as they were
called.
A very large literature has since developed with-
in conservation biology on the concepts of species.
ESUs, and other recently introduced terms such as
"management units." Systematists have pointed out
that ESUs have broad equivalence to the species
units identified by the phylogenetic species concept
and that use of the PSC
conservation community not currently satisfied by
the BSC (Cracraft, 1991, 1997; Vogler & Desalle.
1994; Barrowclough & Flesness, 1996). Other con-
servation biologists, notably those having a popu-
meels the needs of the
lation genetic approach to the problem, have con-
tinued to support the BSC and seek ways to refine
theoretical and practical ue Be to the ESU
concept (e.g.. Moritz, 1994a, b. 1995)
There is, however, a — ii crucial argu-
ment often left out of these debates: if conservation
biology is to be a science that relies on a rigorous
description of diversity, then it should adopt the
anguage and conventions of systematic biology and
(Cracraft. 1997: Wheeler.
Although systematists may argue vehement-
taxonomic practice
1997).
ly over which species concept is best. they agree
on many issues of formal taxonomy—that species-
that
names are tied to type specimens, that there are
level taxa have formal Latin names, those
standard rules of nomenclature (the international
codes) so that scientific names can be organized
and managed over time, and that there must be
voucher specimens to document taxonomic deci-
sions and descriptions. None of this is found
concepts such as ESUs or management units.
Conservation biology should therefore abandon
such concepts as evolutionarily significant units.
ESUs are not a substitute for formal taxonomy. Un-
like formally described taxa with their types and
historical continuity in rules of nomenclature, ESUs
cannot have “legal standing.” ESUs cannot, and
should not. be the units we regulate in trade. pro-
tect with legal instruments, or expect to be used by
applied biology for biotechnology. biodiversity in-
formation systems, and many other uses. This is not
the case with taxonomic units (Geist, 1992). Thus.
the power of systematics and taxonomy: despite ar-
guments over the most fundamental units of na-
i agonists share disciplinary
standards that prevent names, and thus the iden-
tified taxa themselves, from devolving into chaos
over long periods of time.
THE SECOND GREAT QUESTION: HOW MANY SPECIES
ARE THERE?
Discovering and describing Earth’s taxonomic di-
versity is the starting point for all biological knowl-
edge. Because of its scope and complexity, however,
knowing how many species inhabit Earth’s ecosys-
tems is one of the megascience questions of biology.
While it is generally accepted that around 1.5-1.7
million species have been discovered and de-
scribed, estimates of unknown diversity range any-
where from 10 to 100 million species, with 13—20
million being the most frequently seen number
(e.g.. Stork, 1999). So the answer to this megasci-
ence question is: We don't know!
But we must find out. When one realizes that the
use of biodiversity drives the world economy and
this has come from knowledge of about 1.5-1.7 mil-
lion species—probably less than 10% of all species
on planet Earth—it is clear that abundant new ben-
efits will flow from newly documented. diversity.
Among these benefits of inventories identified by
SA2000. one could include:
132
Annals of the
Missouri Botanical Garden
* they document patterns of diversity across habi-
tats and ecosystems
* they provide baseline data for monitoring activi-
ties
they identify areas of endemism and regions in
need of conservation and protection
* they discover new species having economic and
social value to societies (pharmacological, agri-
cultural, fisheries, biotechnological)
they provide the baseline data for implementing
an ecosystem approach to conservation and sus-
tainable development
* they support and promote tourism.
We are disastrously ignorant of the natural world.
That this is so was highlighted in the 45th Annual
Systematics Symposium of the Missouri Botanical
Garden (Richardson, 2000), which summarized the
need for, and challenges to, systematic inventory
( 2000; Lundberg et al., 2000)
and also highlighted that we are in a new age of
discovery (Donoghue & Alverson, 2000). We know
so little that even in the most well-known groups
.g., Prance et al.,
~
such as birds and mammals, new species are being
.. Giao et al., 1998; Be-
1999; MacKinnon, 2000). The
most outrageous and spectacular example of this is
probably the description of six new species of le-
murs (in three separate genera) from Madagascar,
all published in the year 2000 (250 years post-
Linnaeus) in a
discovered each year (e.g
resford & Cracraft,
of the International
Journal of Primatology (Thalmann & Geissmann,
2000; Groves, 2000; Rasoloarison et al., 2000).
single issue
THE THIRD GREAT QUESTION: WHAT
OF LIFE?
IS THE TREE
Reconstructing the Tree of Life (TOL) is the third
great question of systematic biology. This is also a
megascience question for systematics for several
well-known reasons. First, given N taxa, there are
The con-
sequence is that as the solution to the second great
— ] nodes that need to be resolved.
question. of systematics—How many species are
there?—plays out, the domain of this third question
gets larger and larger. A second major reason re-
lates to the first: as the number of taxa in the tree
increases, it gets more and more difficult to obtain
sufficient data on these taxa, and the computational
complexity of finding an objective answer also be-
comes astronomical, Compared to deciphering the
POL, the determination of the sequence of the hu-
man genome, which for all intents and purposes
was solved by a single, relatively small corporation,
turned out to be a rather simple problem (it just
required a little money and coordination among sci-
entists). In contrast, as a scientific problem, resolv-
ing the TOL is much more comparable in complex-
ity, say,
genome,
knowing all the genes in the human
how they function, and how their gene
products interact to form a blueprint for develop-
ment.
So the answer to the third question is, of course:
we don't know. As an index to the vastness of the
problem, we can estimate that there are right now
about 1.7 million nodes on the TOL, reflecting the
number of species that have been discovered and
described. Yet, where do we stand presently? No
one knows for sure, but roughly—very roughly—
perhaps 50,000 to 60,000 species are represented
on one kind of tree or another. That is a pure guess
inasmuch as there is no comprehensive database of
trees. The only repository approaching what is
needed is TreeBASE [TreeBASE], which has ap-
proximately 12.000 taxa, but the sample is highly
biased in being mostly botanical.
Assuming 50,000 taxa already placed on one or
more trees, it is fair to say the position of most of
those is poorly supported by character evidence.
While it is perfectly accurate to say our knowledge
of the TOL is growing very rapidly, as measured by
the numbers of phylogenetic papers being pub-
ished, it is equally accurate to say a large per-
centage of the nodes on those trees have relatively
little support. this,
among the most important being poor taxon and
character sampling, poor choice of character sys-
There are many reasons for
tem, and ambiguities in the methods used to ana-
lyze the data. Moreover, because taxon sampling is
generally poor in most published phylogenetic
studies, it is not at all clear how the different results
can be linked with one another to form a general
view of the TOL, a “supertree of life
” if you will.
The most remarkable observation is that our un-
derstanding of the TOL—at least in terms of the
50,000 taxa just mentioned
is a product mostly of
the last decade. Modern phylogenetics is only about
30—40 years old, and serious
much older.
“tree thinking” not
The rise of “evolutionary systematics”
in the 1930s and 1940s, with its emphasis on a
population biology/genetic approach toward the his-
tory of life, slowed the discovery of the TOL be-
cause it was largely assumed that if ancestors could
not be found in the fossil record, there was little
hope of understanding phylogeny.
If that seems a misrepresentation of history, one
only has to examine the content of the major sys-
tematics journals (for example, Systematic Zoology)
prior to 1960 to see that depicting relationships as
trees was not of major importance. There was re-
markably little “tree thinking" prior to the intro-
Volume 89, Number 2
Cracraft
2002 Systematic Biology
duction of numerical taxonomy, a discipline that
created trees, but was ambivalent in its interpre-
tation of them. Many proponents of this approach
saw their trees as purely representational of phe-
netic similarity, not history; others hoped these di-
agrams might reflect some trace of history. The im-
portant point here is that, compared to evolutionary
systematics, numerical taxonomy developed repeat-
able methods that produced trees.
Both evolutionary systematics and numerical tax-
onomy were eclipsed by Willi Hennig’s conceptual
and methodological development of phylogenetic
systematics. or cladistics. The broad adoption of
cladisties formalized tree-thinking in terms of phy-
logenetic relationships and history. Also, the nu-
merical methods that were rapidly introduced
brought a much needed objectivity, both philosoph-
ical and analytical, to the study of phylogeny. The
explosion of phylogenetic knowledge over the last
decade has resulted just as much from the concep-
tual and analytical revolution of the 1960s to 1980s
as it has by the introduction of efficient methods to
gather new kinds of data, especially those from mo-
lecular sequences.
Why phylogenetics matters
These advances in phylogenetic theory and
have revolutionized systematic and
methodology
comparative biology, and the transformation of sys-
tematies into a truly historical science could not
have come at a better time. Society is desperate for
knowledge about phylogeny. While many systema-
tists still see an understanding of phylogeny as a
goal in itself, numerous segments of society are
looking to phylogenetics to solve entirely new kinds
of problems. Consider the following examples:
l. Tracing disease transmission
The first application of phylogenetic analysis to
examine disease transmission employed parsimony
analysis to investigate whether a Florida dentist.
discovered to be HIV-positive and who had con-
tracted AIDS, had transmitted the infection to any
of his patients (Ou et al., 1992). These investigators
chose the HIV envelope (env) gene because of its
high variability and compared sequences from the
dentist to those of HIV-positive patients and HIV-
positive nonpatients as controls. Phylogenetic anal-
ysis showed clear patterns of genetic relationships
between the dentist and at least five of his patients
who had no identifiable behavioral risks to contract
HIV infection (Fig. 1;
of phylogenetic methods in analyzing HIV evolu-
1999).
tion. see Holmes et al., 1996, and Crandall,
for discussion of other uses
— Dentist-x
Patient D-x 4%
Patient О-у
Figure l. P орава netic tree of HIV-1 епу V3 sequenc-
es from a HIV posi Florida dentist dee his patients
A—G (x and y ce efe r to divergent clones, LC refers
Ju et al., 1992, for details). These
with the hypothesis that the dentist
section. This was
to local controls: see (
results were consiste nt
was the source of the patients’ iw infe
the first use of phylogenetic analysis to examine
о Reprinted with permission from Ou et al.
65, figure 1. Copyright 1992
Advancement of Science;
disease
~
(19 Science 256: 11
— in Association for the
http://www.sciencemag.org.
Tracking the spread of “emergent” diseases
2 | g
Phylogenetics is playing an increasing role in the
medical sciences, especially in identifying disease
agents that spread from one region of the globe to
another. DNA sequences from disease entities can
be rapidly obtained and compared to sequences
housed in databases such as [GenBank]. The sum-
mer of 1999 in the New York City area brought a
strange confluence of events. A number of people
were stricken with an encephalitis that had the eti-
ology of a flavivirus. At the same time large birds,
including American crows (Corvus brachyrhynchos).
were turning up dead in unusual numbers in the
wild and in local zoos. Viral particles were even-
tually isolated and their polyprotein nucleotide se-
quence determined. Phylogenetic comparison with
other sequences identified the newly emergent dis-
ease as being related to West Nile Virus (Fig. 2;
Lanciotti et al.. 1999), which circulates between
birds and mosquitos and from the latter into hu-
mans. Similar strains in the Mediterranean region
and Middle East were also associated with in-
creased avian mortality. The virus has now spread
134
Annals of the
Missouri Botanical Garden
WN-Romania 1996 H
distance
0.045
Lineage 1
N-M
—— 1988
N-Kenya
un cagas 1986
-Uga
Lineage 2
Figure 2.
origin of the West Nile-
New York City region in 19
These results indicate the New York stra
lated to strains ina central and northern Africa, the Mid-
dle East, and eastern Europe and was derived from that
region. Reprinted with. — from Lanciotti et al.
19 Science 286: 2333, figure 2. Copyright 1999
American Association з the Advane етеп of Science;
http://www.sciencemag.org.
Phylogene i analysis was used to trace the
like virus that broke out in о
J9 (Lanciotti et al., 199
un is closely re-
well beyond the New York region and also into non-
human mammalian hosts.
emergent diseases
Discovery of “new”
In addition to tracking diseases from one region
to another, phylogenetic analysis is being used to
discover new disease entities. In late 1998 and ear-
ly 1999 a new mosquito-borne virus, called Nipah,
emerged in Malaysia (Chua et al., 2000). Using pigs
as a vertebrate host the virus jumped to humans,
causing symptoms that first suggested Japanese en-
cephalitis. Eventually 265 cases were reported and
105 people died from severe nervous system pa-
thology. To control the epidemic, over a million pigs
had to be slaughtered.
role in
Phylogenetic analysis played a major
helping to characterize Nipah virus (Chua et al..
2000). Comparative sequences were obtained from
the nucleoprotein (N) gene and compared to other
members of the subfamily Paramyxovirinae. The re-
sulting tree demonstrated the relationship of Nipah
to another recently discovered virus, Hendra virus,
and the sequence differences indicated they were
distinct (Fig. :
4. Monitoring and predicting viral host switching
Karposi's sarcoma virus is endemic to central Af-
rica and has associated with ita rhadinovirus, Kar-
pos's-sarcoma-associated herpesvirus (KSHV), Un-
til recently rhadinoviruses (y,-herpesviruses) were
found in various Old and New World monkeys but
not human’s closest relatives, the great apes. After
determining sequences of herpesvirus DNA poly-
merase taken from wild Pan troglodytes and Gorilla
gorilla from Cameroon and Gabon, Lacoste et al.
(2000) reported the discovery of new strains of
these viruses. When those sequences were com-
pared to others already known using a phylogenetic
analysis, Lacoste et al. (2000) showed that these
new viruses are closely related to KSHV (Fig. 4).
The phylogenetic closeness of these new herpesvi-
ruses and KSHV raises the potential for host
switching into humans as they hunt and consume
great apes for food. Phylogenetic analysis contrib-
utes importantly to identifying and monitoring this
new health threat.
5. Genomics, development, gene expression, and
disease
Phylogenetic thinking and methodologies аге
taking hold in the fields of genomics and molecular
medicine (e.g., Eisen, 1998; Pollock et al., 2000)
and promise not only to increase our knowledge of
the relationships of organisms but also surely will
lead to insights into understanding and predicting
gene structure and function. Developmental biolo-
gists have long acknowledged the predictive and
explanatory power of phylogenetic relationships in
reconstructing the historical pathways of develop-
ment (reviewed in Raff, 1996), and the rapidly ex-
panding field of evolutionary development (“evo-
devo”) will, reciprocally, result in major new
advances in understanding developmental mecha-
nisms and will inject new character systems into
systematics that will inform phylogenetic relation-
ships of major organismal groups.
6. Identification of invasive species
The transport of alien species is a major global
environmental problem. The United States alone
Volume 89, Number 2
2002
Cracraft
Systematic Biology
135
— 100 nt substitutions
Genus Respirovirus
Genus Morbillivirus
Genus Rubulavirus
Figure 3. Phylogenetic analysis of the nue ‘leoprotein (N) gene of the so-c called de virus that broke out in Malaysia
in 1998 “and 1999 (Chua et al.. 2000) showed this
paramvxov irus,
2000 American Association for the Advancement of Science:
harbors about 50.000 invasive species at a loss of
nearly $137 billion annually to control and mitigate
their effects (Pimentel et al., 2000: Wolfenbarger &
Phifer, 2000). Identifying potential exotic species
is a major priority and a first-line of defense against
them. In 1984 a tropical marine green alga (Cau-
lerpa taxifolia) escaped from an aquarium and in-
vaded the Mediterranean Sea. This particular strain
proved remarkably hardy and competitive, and
spread rapidly to devastate populations of native
species (Jousson et al., 2000). The species was re-
cently discovered several locations along the
California coast. Jousson et al. (2000) posed the
question whether the California populations of Cau-
lerpa could be identified as an invasion of the Med-
iterranean strain, representing a potentially serious
threat to coastal ecosystems. Comparing DNA se-
quences from the internal transcribed spacer of ri-
bosomal DNA from multiple populations, phyloge-
from
netic analysis united 11 of 12
samples in California with those from the Mediter-
sequences
ranean: one sequence clustered with a Red Sea/
IndoPacific clade (Fig. 5). It was concluded that an
immediate eradication program was warranted.
7. Discovery of microbial diversity
Phylogenetic analysis of DNA sequences has be-
come a major tool in the discovery of new micro-
organisms, especially bacteria. Because most «
these organisms cannot be cultured, microbiologists
have turned to molecular probes for inventory and
s new emergent virus
Hendra virus. Reprinted with permission from € hua etal. ( 2000). — 288:
ated to another recently discovered
1432. figure 4. Copyright
Was с lose ly
: http://www.sciencemag.org.
identification (Pace, 1997). Typically using probes
for rRNA genes, the sequences are compared
those databases by various phylogenetic tech-
niques. Phylogenetic methods have thus opened up
entirely new approaches to understanding the mi-
erobial diversity of extreme environments (Horiko-
shi & Tsujii, 1999) and have led to a greater un-
derstanding of the distribution of microbial life
forms. It is now appreciated, for example, that ar-
chaebacteria are not only found in extreme envi-
ronments such as hydrothermal vents and hot
springs but are much more widespread than pre-
viously thought, including a variety of coastal and
open ocean habitats (DeLong, 1992).
THE FOURTH GREAT QUESTION: WHAT HAS BEEN
ГНЕ HISTORY OF CHARACTER TRANSFORMATION?
The proposition that the history of character
transformation might be considered a great ques-
tion of systematics may strike some as a bil strange.
but reflection will confirm that all we know about
the evolution of form and function derives from how
character change is interpreted relative to a given
which
танк have been two main ways in
sformations are studied. First, in gen-
tree.
character tra
eraling esa hypotheses under maximum
parsimony, characters used build the tree are
optimized on it thus allowing inferences about their
transformation across the tree. A second method
has been to take some tree as given and then op-
timize. or plot, characters on it. There is a growing
136 Annals of the
Missouri Botanical Garden
100 75 TM. mulatta ow
100 М. жассан
00 М. nemestrina
o6 ChRV2
EBV-B958 « y1 J
MCMV
89
HCMV
80 p-Herpesvirus
an
o ives
угу 7]
GHV2
Figure 4. A phylogenetic tree based on sequences of the DNA polymerase gene for several herpesviruses newly
discovered in the chimpanzee (Pan troglodytes) and lowland gorilla (Gorilla gorilla) in Cameroon and Gabon (Lacoste
et al., 2000). These results indicated these new viruses were closely related to the Karposi's-sarcoma-associated her-
pesvirus (KSHV) found in humans. This suggests the possibility that the new herpesviruses might be transmitted to
humans since chimps and gorillas are frequently used for food; thus, phylogenetic analysis can be used to predict
possible outbreaks as well as help establish a monitoring program for new infection. Tree modified and reprinted with
permission from Lacoste et al. (2000), Nature 407: 151—152, figure 1. Copyright 2000 Nature; http://www.nature.
сот/.
literature arguing and demonstrating empirically no phylogenetically relevant information, which is
that the first approach is to be preferred, primarily seldom true.
because the characters of interest are often deter- Nevertheless, the second approach of plotting or-
minative with respect to choice of most parsimo- — ganismal attributes on a phylogenetic tree will re-
nious tree. The second approach tacitly presumes main popular and exceedingly important. Indeed,
that the characters being examined have little or the main benefit that nonsystematists gain from a
Volume 89, Number 2
2002
Cracraft 137
Systematic Biology
Mediterranean
California
Australia
i a Australia
Indonesia
New Caledonia
Red Sea
Australia
Caribbean
Red Sea
Figure 5.
the place of origin of a recently discovered invasive alga
( Caulerpa taxifolia) along the California coast (Jousson et
al., 2000). the internal tran-
scribed spacer of ribosomal I l
strains, it was found that the Pa ar strain clustered
with those from the highly invasive strain th
considerable damage in the Mediterranean. This suggest-
ed the need for an immediate eradication program. Re-
ке with жн from Jousson et al. (2000). Nature
—158. figure Ib. Copyright 2000 Nature: http://
a el
Phylogene ‘tic analysis was used to ascertain
Comparing se — es from
>
£e
=
>
=
ys
—
—
—
fuller understanding of the Tree of Life is that it
helps them understand the history of characters and
make predictions about taxa for which those char-
acters are yet unknown. Significantly, many of the
users of phylogenetic information are themselves
contributing data about character systems that will
potentially inform our understanding of relation-
ships.
Literally hundreds of papers have used phylog-
enies to interpret nonsystematic data, and there is
no question that this has led to numerous insights
into the history of many character systems in be-
havior, ecology, physiology, and other sciences (e.g..
Harvey et al.. 1996). More important. perhaps. will
be the power of character transformation analysis
in applied biology. It has been used. for example.
to guide the search for new pharmaceuticals or bio-
chemical products (see Monson, 1996, for review).
Peterson
—
understand the history of gene regulation
& Davidson, 2000) and body complexity (Graham
et al.. 2000: Cameron et al., 2000). among others.
Answers to the fourth great question will take on
more importance as phylogenetic knowledge and
comparative databases expand and become more
available.
EARTHS
THE FIFTH GREAT QUESTION: WHERE ARE
SPECIES DISTRIBUTED?
This is the most fundamental question of bioge-
ography and any answer will only be as good as the
inventories on which it is based. Thus, knowledge
of distributions relies on the presence of georefer-
enced specimens housed in the world’s natural his-
tory collections. Ultimately, the characterization of
Earth's habitats and ecosystems depends on these
data, as does the ability to manage and conserve
biodiversity.
The practical importance of having knowledge of
species’ distributions is acknowledged by the desire
of nations, and
intergovernmental organizations,
nongovernmental conservation organizations 10
have specimens in natural history collections da-
tabased and made freely available on the internet.
The development and availability of geographical
information systems (GIS) and other software pro-
erams for mapping diversity are also increasing the
value of digitally captured specimen data (e.g...
Funk. 1997). Although such data are increasingly
coming online, most of the world’s collections are
not databased. This has led some governments,
most notably Mexico, to accelerate data acquisition
on their own. The value of their efforts has been
well documented (e.g... Bojorquez-Tapia et al.,
1994: Soberón et al., 2000) and recognized the
as evidenced by the formation of the
world over,
Global Biodiversity Information Facility [GBIF].
Distributional information for individual species
leads to the search for patterns of diversity at dif-
ferent scales. Most importantly, this builds knowl-
edge about areas of endemism, and discovering ar-
eas with high numbers of endemic taxa (sometimes
referred to as “hotspots”) is widely considered crit-
ical for se tting Cc — к 5 (F orey e ta
1994: Nielsen & : Nielsen, 1999). But
distributional data * a E larger significance
for society than simply that associated with conser-
vation. Distributional information tied to specimens
underlies drug discovery, ecotourism, trade in nat-
ural resources. pest control, control of invasive spe-
cies, crop improvement using the genetic diversity
of wild relatives. and analysis of global change.
among many other applications.
THE SIXTH GREAT QUESTION: HOW HAVE SPECIES
DISTRIBUTIONS CHANGED OVER TIME?
This great question of systematics, and the sec-
ond pertaining to biogeography. can be looked at in
two ways. First, in the here and now: The vast ma-
jority of research and activity in
in ecological biogeography, in which ecologists at-
“biogeography” is
tempt to understand and explain why organisms are
distributed where they are, how those distributions
are tied to autecologies, and so forth. Yet, without
accurate taxonomic descriptions and georeferenced
138
Annals of the
Missouri Botanical Garden
data vouchered by specimens, the quality of the
ecology itself will suffer. A component of this re-
search also looks to the future. Human activities
are transforming the biosphere and there is interest
in predicting how this impact will affect the distri-
butions of organisms and, downstream, societal
well-being, especially for critical activities such as
agriculture.
The second way of thinking about changes
distributions is through the eyes of the historical
biogeographer. Following the introduction and ap-
plication of cladistic methods to distributional
problems (Rosen, 1978; Nelson & Platnick, 1981),
interest in the biogeographic history of taxa and
areas of endemism expanded significantly. As was
realized early on, the key to understanding the his-
tory of biotas is through an analysis of the history
1986). Areas
of endemism are evidence that components of bi-
of areas of endemism (e.g.. Cracraft,
olas (species) have become isolated and differen-
tiated, and nested areas of endemism are evidence
that biotas have expanded through dispersion to be-
come
more or less cosmopolitan, only to divide
again. Yet, reconstructing this history has not been
easy. It has become clear that most of these histor-
ical patterns of distribution are so complex that no
single method of analysis—at least not one that is
currently known—is capable of giving a completely
satisfying resolution. Each method of biogeographic
analysis, it appears, has various shortcomings in
how it handles widespread taxa, redundant distri-
butions, and missing taxa and areas, in addition to
the fact that the history of areas itself may not al-
ways reflect a hierarchical (branching) pattern.
THE SEVENTH GREAT QUESTION: HOW IS
PHYLOGENETIC HISTORY PREDICTIVE?
The third mission of systematics, as identified by
SA2000, was to create an efficient, and predictive,
systematic information system. This included da-
tabasing specimens and making the information
widely available, linking to other biodiversity and
biological databases, and building informatics ca-
pacity
to utilize biological—and systematics—
knowledge globally. The predictive element was en-
visioned as coming from using phylogenetic
classifications to guide searches for information.
and thus reflect the hierarchical relationships of
ife.
The expectation that closely related taxa. share
similarities not shared with more distant taxa is the
foundation for comparative biology. An information
system that is queried using the hierarchical rela-
tionships of life can be termed phyloinformatics (the
power of “phyloinformatics” was noted in an NSF-
sponsored workshop [Tree of Life| and as well as
by Edwards et al. (2000). Thus, the ability to search
multiple databases using the nodes of a phyloge-
netic tree may be the single most important contri-
bution of systematics to conservation and sustain-
able use of biodiversity. Searches that query across
databases of various kinds from the perspective of
phylogenetic groupings would therefore have im-
mense predictive power because the resulting data
can be expected to reflect attributes shared by, or
potentially shared by, those groups.
DISCUSSION: THE FUTURE OF SYSTEMATICS AND
Irs RELEVANCE
DIVERSITY
A safe prediction is that debates over the species
question will continue. However, I think there is
far more actual agreement among practicing sys-
tematists than could be concluded by recent papers
supporting the BSC. This comment is not to dis-
parage these viewpoints. Rather, it is to reaffirm the
aput of others (Nixon & Wheeler, 1990; Wheel-
„ 1997) that taxa, and taxic limits, as well as the
— governing names, are the primary domain of
practicing systematists. And adjudic ating taxic lim-
its, moreover, is very much a practical process: as
taxa are analyzed with the purpose of understand-
ing patterns of character variation, geographic dis-
tributions and endemism, relationships, and bio-
geographic history, we will more and more need a
species concept that looks at basal taxonomic units.
That is the trajectory of research by young system-
atists doing these kinds of studies, and that trend
will certainly continue.
But fighting over species limits is not the frontier
of the theme of diversity—rather it is discovering
and describing the other 95+% of life. Society
needs to know what species share the planet with
us, and the urgency has never been greater. Inven-
torying life's diversity is a megascience question
because of its enormous intellectual and technical
challenges. Even understanding how to inventory a
single taxonomic group in a circumscribed region
is a difficult problem (e.g., Coddington et al., 1991),
let alone thinking about what approaches and de-
signs of inventories might be appropriate for inves-
tigations that are global in scope (Wheeler, 1995a;
Wheeler & Cracraft, 1996). A global inventory of
life will therefore require meticulous long-term
planning and capacity building in infrastructure
and human resources, and it will be expensive. But
it should be a shared expense of governments, in-
tergovernmental institutions, and the private sector.
Volume 89, Number 2 Cracraft 139
2002 Systematic Biology
Indeed, the private sector is getting involved, and molecules alone are yielding the "true" tree of life.
with a larger imagination than currently witnessed
within most. governmental and. intergovernmental
circles. Thus, a consortium of technology leaders
with a track record in futurist thinking have rec-
ognized the need for a global inventory of species
and are beginning to organize a long-term effort to
see the job accomplished [АП Species Inventory].
Because of this brilliant and far-sighted effort. in-
ventory should take on new life!
PHYLOGENETIC HISTORY
In the previous section so much space was de-
voted to phylogenetics and the Tree of Life because
it is often not widely realized, even in the svstem-
atic community, how crucially important that
branch of systematics is becoming. Phylogenetic
knowledge is exploding and the rate at which this
is happening will not diminish for quite some time.
Problems that undermine the quality of phyloge-
netic research are commonplace and unfortunately
will probably continue to persist. Many non-system-
atically trained biologists, almost all using molec-
ular data, have paid scant attention to taxonomic
documentation, nomenclatural issues, or lo proper
vouchering of sequences by reference to specimen
data. Failure to heed these problems will lead to
errors and even to potentially dangerous conclu-
sions in fields like human health (Ruedas et al..
2000). Systematists need to work with editors and
editorial boards of journals, and with data deposi-
tories like GenBank, to improve this situation.
There is a tacit assumption on the part of many
researchers building trees from molecular data that
their results are inherently superior to those relying
on morphological data. Results are presented, with
little or no discussion, and often outright dismissal
or total ignorance, of prior morphological studies.
This attitude developed soon after molecular stud-
ies were introduced (the “molecules versus mor-
phology” debate; e.g.. Patterson, 1987). and as |
said, it continues.
This conflict is likely to increase as high
through-put methods of genomics find. their way
into laboratories doing comparative phylogenetics.
The vast majority of current studies utilize small
taxon samples and molecular data samples. leading
to a significant accumulation of questionable results
because of sampling artifacts. But this situation ts
changing. As methods improve and more resources
are allocated to molecular work, more studies are
significantly expanding the sizes of matrices i
(>)
terms of both taxa and characters. Thus, | predict
there will be an increasing tendency to think that
But this would be a big mistake. All major groups
of living taxa have many long-branched lineages
that are monotypic or have relatively few closely
related species, and these will in all likelihood con-
tinue to confound analysis of deep-branch relation-
ships. To resolve those relationships satisfactorily,
it will almost certainly take the addition of mor-
phological characters, especially those from fossil
taxa. The relationships of many higher taxa, such
as mammals and birds and many others, have been
exceedingly difficult to resolve because of the prob-
lems just mentioned, and our future hope of un-
derstanding the history of life will lie in using all
available data.
Applied phylogenetics, as it might be termed, is
an easily identifiable wave of the future. One of the
most remarkable signs of the vitality of phyloge-
netics is the expansion of its use into human health,
developmental biology, forensics, natural resource
management, and other areas. A large number of
phylogenetically oriented young systematists are
seeking careers in the biotechnology and genomics
industries where understanding of phylogenetic
methods, and a comparative approach to problem-
solving, are needed. This trend will continue for a
long time to come.
BIOGEOGRAPHY
Given the current state of knowledge of global
biodiversity (5-10% known), and given that most
of the species (insects) already described are
known from only a handful of localities, it is fair to
conclude that we have very imprecise knowledge
of the distributions of Earth’s species (although sin-
ele localities could be interpreted as being fairly
precise). The task ahead is daunting because vir-
tually all of Earth's habitats and ecosystems have
been incompletely inventoried, even for well-known
eroups such as birds (e.g.. Peterson et al.. 1998
and mammals (Patterson, 2000). Although an in-
crease in our knowledge about distributions will ul-
timately be linked to the intensity of inventory ef-
forts, electronically capturing and correlating what
we have already collected will surely increase our
understanding of distributions and patterns of en-
demism.
Biodiversity cannot be managed or conserved
without distributional information, thus the unde-
niable importance of databasing the world’s natural
history collections. Moreover, as global climate
change accelerates, and as the anthropogenic con-
version of habitats continues, predictions about the
140
Annals of the
Missouri Botanical Garden
Phyloinformatics: A Conceptual Framework
Databases
Specimens/Collections
| Genetic/Genomic| | Ecological | | Geological
Associated biological data
f, n А dat
Geo
UCO
| Developmental | | Land use
| | | Geographic |
Associated literature, etc
| Human health | | Economic/trade | | Agricultural |
N
А
Node-based Search
D «“— Species/Taxon
Nodes/hierarchy —» 2
Fig
re ‘Lationships to drive queries among multiple database
ure 6
d query for information for all species/taxa (/
on the [Tree of Life | held at Yale University in July 2
consequences of these changes become more and
more important. Many of those predictions will flow
from the use of historical distributional information
linked to specimens.
Biogeography is one of the great frontiers of
systematic research. Factors such as widespread,
redundant, and missing distributional data hin-
der understanding of history, but these
problems themselves suggest that the field is still
wide open to theoretical,
biotic
methodological, and
empirical research. Knowledge of biogeographic
history is so central to understanding patterns
and processes of biological diversification,
cluding speciation,
in-
as well to how biotas
evolve over time, that it will continue to be a core
as
The concept of a phyloinformatic search strategy uses the hierarchy implied by knowledge of phylogenetic
tead of unde rtaking searches one species at a time, nodes
earches for ялы тее by related species. Here, searches on node (or line 'age)
с ) — that node. This figure was first used in a NSF works shop
area of research for many years to come. There
have been very few studies of biotic history that
integrate patterns of relationships among areas of
endemism with information from paleontology,
paleoclimatology, and paleogeography. There is
an unfortunate disjunct, on the one hand, be-
tween paleontologists who study diversification
over time and rarely concern themselves with
present-day patterns and processes such as spe-
ciation, and those neontologists who look at pat-
terns of historical biogeography using distribu-
tions of the Recent but who rarely
incorporate paleontological data on changing di-
versity. A bridge will have to be built before we
have a satisfactory picture of biotic history.
biota
Volume 89, Number 2
2002
Cracraft
Systematic Biology
PHY LOINFORMATICS
| predict that phylogenetics will have its greatest
societal impact by empowering and enriching the
search for information and data associations across
many different kinds of biological and systematic
databases. Information flow will make peoples’ lives
better. That is what phyloinformatics can, and will,
do.
While queries of biodiversity databases will al-
ways make use of species” names as pointers lo
information (Bisby, 2000).
queries can expand and integrate searches and in-
the use of node-based
formation to another order of magnitude (Fig. 6).
classifications will facilitate. а
Phylogenetic new
way of gathering biological information and linking
it to other nonbiological databases. The establish-
ment of the Global Biodiversity Information Facility
(СВІ: see 2000) on 1 March
2001 holds the key to making biodiversity infor-
Edwards et al..
mation readily accessible to all. Phyloinformatic
queries will expand the potential of GBIF in ways
not previously imagined.
The predictive power of phyloinformatics. based
as it is on an understanding of the relationships
among Earth’s species, argues persuasively for dis-
covering all branches of the Tree of Life as rapidly
as possible.
Literature Cited
Agenda Systematik 2000. 1996. Agenda Sy stematik 2000
Erschlie ssung der Biosphire. Kleine Senckenberg-
Reihe Nr. 22: 1-55.
American Museum of Natural History.
Taxonomy Initia Using
Jeet Country iad Renomi Nee
1999, The Global
stematic Inventories to
. Center for Biodi-
cuin an Мил seum of Natural
ps e also (http://researc ^. amnh.org/
i.html)]
ve rsity and C —
\
Anonymous. 1991. Sy ste matics s Agenda 2000: Integrating
biologic al diversity and societal needs. Syst. Zool. 10:
520-52:
Avise, J. . M. Ball. 1990. Principles of genealogic al
cone — e in species conce pts and biulosic al taxon-
Д у. Evol. Biol. 7: 45-67.
hus P. 1999, i
ia lip service ar action? Courier Forschunginst. Set
enbere 215:
—From Convention
pm
. pois: ethnobotany for the new
ا Ann. Missot
i Bot. Gard. 83: 58—00.
n Fles 1996. Species.
The problem of units of man-
-25
sness.
races:
agement in conservation. Pp. 247 4 in D. G. Klei-
man, M. Alen, V. >. Lumpkin & H.
Harris (e ditam. Wild Mammals in € ap tivity: Principles
and Тес — s. Univ.
Beresford, Cracraft. 1999, Speaistion in African
orest — (Stindroruis) Species limits, phylogenetic
Thompson, 5
Chicago Press, Chicz ago.
re а and molecular biogeography. Amer. Mus.
22.
Novit. 3270: 1—
Bisby. F. А. 2000. The quiet т — in-
formatics and the internet. Sci » 289: 23
ке S. р. Cutler. 199 o. dr henna
The Challenge for Europe. Linn. Soc. Occas.
London.
wa No. 1, |
Bock. W. J. 1987. рене ies concepts, speciation, mac-
evolution Pp. 31- 1D. Iwatsuki, P. Raven №,
J. Bo k fedi Moden "n о of Species. Ü niv. To-
kvo Press. Tok »
Bojorquez- "Tu a. P. Balvanera & A. D. Cuaron. 1991.
iological inventories and computer databases: Their
role in environmental assessments. Environm. Mana-
vem. 18: 545-551.
d p. a D cLennan, J. M.
‚ 1995. Systematics, ecology.
Carpenter, 5. 6.
a id De — "Bias ience 15 687-095.
Cameron, C. B., J. R. Garey & B. J. Swalla. 2000. Evo-
lution of the chordate body plan: New insights from
phylogenetic analyses of deuterostome phyla. Proc.
Natl. Acad. Sei. U.S.A. 97: 4409-4474
hua. K. B.. W. J. Bellini, P. A. Rota. B. H. Harcourt. A.
lamin. 5. K. Lam, T. G. Ksiazek, P. E. Rollin, S. R.
aki. W.-J. Shieh, C. S. — D. J. wes JT
Roehrig. B. Eaton, A. R. Gould, J. Olson, Field, P.
Daniels, A. E. Ling, C. J. T Leds our & B.
2000. Nipah А rec ently | emergent
de dh param) xovirus. Science - 288; 432-14.
Claridge. F. 1995. Introducing eat s
2000.
ge
Badio sity & Conservation 4: 451-454,
HA. Dawah 8 Wilson (editors). 1997.
Species: The Units of Diversity. Chapman & Hall, Lon-
don
— J. ^.. E. Griswold,
aranda & S. F. I w d 199].
sampling air ie to estimate ег in tropical
) in E. C. d (editor), The
Unity of E — Biology, J . Dioscorides Press,
Portland. ( regon.
а A. Orr & D. J. Futuyma.
need a new species
Cracraft. J. 1983. Species concepts and speciation anal-
Ornithology 1: 159-187.
36. Origin and t. Men of continental biotas:
D. S. Dávila, E. Pen-
Designing г testing
ecosystems. Pp. 41-6
1988. Do we
190—200.
Coyne, J.
oncept? Syst. Zool. 37:
ysis. Curr.
. 19%
Speciation and historical congruence within the Austra-
lian m E volution 40: 977 ).
1987. Species concepts Ме: the ontology of evo-
‘ition: Biol, : Philos. 2: 329-346.
89a. Speciation and its ontology: Т he empirical
consequences of alternative species concepts for un-
de — patterns Ke processes of differentiation.
—59 in D. Otte & J. A. Endler (editors). Specia-
tion dd its Consequences. Sinauer, Mas-
Sunderland,
sachusetts
198 0b. Species as entities of biological theory.
Pp. 31-52 in M. Ruse (editor), What the Philosophy of
Kluwer, Dordrecht
Systematics, s
Abstract, 5th I Meeting, Soci-
'onsin.
995. The urgency of building global capacity
Biology Is. А
199] ecies concepts, and con-
servation biology.
ety of Conse rvation Biology, Madison. Wis
|
[ог — science. Biodiversity & Conservation 4:
163—475.
1997. Species concepts in systematics and con-
servation ME An — 'al
325—339 in M. A. Claridge, Н. A. Dawah &
Wilson (editors), Species: The т nits s of Diversity. Chap-
man & Hall,
viewpoint. Pp.
M. R.
London.
142
Annals of the
Missouri Botanical Garden
000. Species concepts in theoretical and ap-
plied — A systematic debate with consequences.
Pp п Q. D. Wheeler & R. Meier (editors), Spe-
cies C once m and Phylogene tic Theory: A Debate. Co-
lumbia Univ. Press, New Yor
Crandall, — 1999,
— Hopkins L niv. Press, Balti
295. Systematics aid ibli health. Bio-
14.
>
—
Te ны of HIV.
. 1992. Archaea in costal marine environ-
. Natl. Acad. Sci. U.S.A. 89: 5685-5689.
& W. S. Alverson. 2000. A new age of
-Momon Bot. — 87: 110-126.
. А. Lane & E. S. Nielsen. 2000. Inter-
ope — ol — data —
mation on every desktop. Science 289: 2312-2314.
Eisen, J. A. 1998. Page s: Improving functional
predictions for unc — d
analysis. 2 ıe Res. 3-167
З M. (editor). 1992. The Units of Evolution: Es-
sa) Nature of Species. MIT Press, Boston.
Eshbaugh, W. H. 1995. Systematics Agenda 2000: An 2g
toric ‘al perspective. Biodive sity & Conservation 4: 455
m
-
X =
=
5
=ч
Biodiversity infor-
genes by evolutionary
Mis P. L., C. J. Humphries & R. I. Vane-Wright. 1994,
Syste ipm s and Conservation Evaluation. Oxford Univ.
Press, Oxford.
Funk, V. A. 1997. Using collections data and GIS to ex-
amine biodiversity information levels in Guyana. Pp.
17-128 in К. К. Hoaglund & A. Y. Rossman (editors),
Global Genetic Resources: Access, Ownership, and In-
tellectual Property — баке. "lation Systematics Col-
lections, — D.
Futuyma, D. J. 1998. Evolutionary Biology,
auer, ж a Massachusetts.
Geist, je np species and the law. Nature
Third Ed. Sin-
. "uos, V. V. Dung, E. D. W —
ato E Arctander & J. R. MacKinnon. 1998. De-
sc иса of Muntiacus id ee из, a new species of
muntjac (Artiodactyla: Muntiacidae) from central Viet-
ат, and implications pu conservation.
servation 1: 61-68.
Graham, L. E., M. E. Coo
Animal Con-
& J. S. Busse. 2000. The origin
of plants: Body plan — contributing to а —
evolutionary radiation. Proc. Natl. Acad. Sci. U.S
ii 40.
Groves, P. 2000. The genus Cheirogaleus: Unrecog-
nized bc in dwarf lemurs. Int. J. Primatol. 21:
943—902.
Harvey, P. H., „ Brown, J. M. Smith € S. Nee. 1996.
Ne ^w Uses 8 Ne W — nies. Oxford Univ. Press, Ox-
Holmes, E. C., P. L. Bollyky, S. Nee, A. Rambaut, €
Garnett & P. — 1996. Using phylogenetic trees to
rec т the history of infectious disease epidemics.
Pp. 169-186 in P. Н. Harvey, A. J. L. Brown, J
datus j^ S. Nee (editors), New Uses for New Phyloge-
nies. Oxford Univ. Press, Oxfor
Horikoshi, К. & K. Tsujii (editors). 1999, He ши
in Deep-Sea sta nts. Springer-Verlag
Howard, D. J. & . Berlocher p ms tm Endle 88
ا Species — Speciation. Oxford Univ. Press,
w York.
ы Т. 1995. Down in the woods the y have no names—
BioNET-INTERNATIONAL.
р)
—
o
=
Strengthe ning systematics
in developing countries. Biodiversity & Conservation 4:
501—509.
Jousson, O., J. Pawlowski,
F. Dini, G. Di Guiseppe,
— 200 ). Invasive
co
Z. Zaninetti, F. W. Zechman,
R. Woodfield, A. Millar & A
alga reaches California. Nature
— . B. Martin (editors
м
1993. Ѕресіеѕ,
Species С one vind and Primate Evolution. Plenum
Press, * ork.
Lacoste, Mauc wu — ul, " Lewis, M.-C.
— & A. . 2000. KSHV-like her-
pesviruses in chimps т “gorillas. Nature 407:
152.
Lanciotti, R. S., J. T. Roehrig, V. Ds ubel, J. Smith, M.
Parker, K. Steele, B. Crise, К. E 3. Crabtree,
H. Se therret, R.
Cropp, B. Panigrahy, E
inson, C.
151-
Ostland, B. Schmitt, м. Malk-
J. Weissman, N. Komar, H vage,
nias ا for an — of
«de T.
a Vl ‚ B. Huey, R. K. Monson & R. J. Jensen.
05. ata s and the study of organismal form and
— BioScience 45: 696—704.
— J. G., M. Kottelat, G. R. Smith, M. L. J. Stiass-
& A. C. Gill. 2000. So many fishes, so little time:
(dip overview of recent ichthyological disc overy in con-
tinental waters. Ann. Missouri Bot. Gard. 87: 26-65
Mackinnon, J. 2000. New mammals in the 21st century?
Ann. Missouri Bot. Gard. 87: 63-66
Mayr. E. 1942
ystematics and the Origin of Species.
Columbia — Vork.
Press, New Yor
————. 1963. Animal Species and Evolution. Harvard
Univ. n ress, Cambridge. Massachusetts.
— ———. 1982. A local — and the biological species
Amer. J. Bot. 79; 222-238.
cone n
— ———. 2000. The lone s spec ies concept. Pp
.17-29
үз Wheeler & R. Meier (editors), Specie s Con-
ea and Phylogenetic Theory: A Debate. Columbia
Univ. Press, New York.
McNeely, J. A. 1995. Keep all the pieces: Systematics
2000 and world conservation. Biodiversity & Conser-
9.
e
1995. Systematics, bio-
ч udis 680—680.
K.
. BioScience 45: €
The u use of phylogenetic perspect
in compara — puso and ‘velopment | bi-
ology. Ann. Missouri Bot. Gard. 83: 3-16.
Moritz, C. 1994a. aa of е 'hondrial DNA
analysis i in conservation: A critical review. Molec. Ecol.
: 401411.
. 1994b. Defining rge — ant units”
for conservation. Trends Ecol. Evol. 9: 373-375.
— ———. 1995. Uses of molecular ph plis ا for ¢ conser-
vation. Philos. Trans. Roy. q^ 349B: 113-118
Nelson, G. & N. I. Platnick. 1 'stematics ind Bio-
geography: C ladistics and Vic 'ariance. Columbia Univ.
Press, New Yo
1999 . Systematics and conservation. Pp.
13 in W. Ponder € D. Lunney (editors), The Oth-
er 99%, The em is Biodiversity of Inverte-
brates. Trans. Roy. Zoo . New South Wales, Mos-
man, New Pis Wales.
& з. West. 1994. Biodiversity research ge bi-
— ‘al ea Transfer of information. Pp.
121 in P. L. Forey
01—
„С. J. Humphries & R. I. Mte
Volume 89, Number 2
2002
Cracraft 143
Systematic Biology
(editors), Systematics and Conservation Evaluation.
Clarendon Pres 288, E ford.
Nixon, K. €
2 poten species concept. Cladistics
oon T. T Н.. Jr. 1996. Importance of systematics to public
health: Tic : microbes, anc
Bot. Gard. 83: 37—40.
disease. Ann. Missouri
Ou, C.-Y.. C. ri Ciesielski. G. Myers, C. I. Bandea. C.-
C. Luo, B. T. M. Korber, J. 1 . Mullins, G. Schochetman.
R. L. Berkelman
L. A. N. Fconomou, + J. Witte, L. J.
Furman, G. A. Satten, Mac s, J. W. E urran &
І Jaffe, Laboratory ов iain. group, Epidemio-
2
logic Investigation group. 1992. Molecular epidemiolo-
ey of HIV Ои in a die practice.
256: 1165-1
Pace, N. R. ard
and the — re.
Patterson, B. D.
ery of new Noti ‘al mammals. Diversity and Distri-
butions 6:
Patterson, €
Science
n e scular view of microbial diversity
ence 276: (34—740.
ce rns and trends in the discov-
15-151.
— 1987. Molecules and ub d in
E лю Conflict or Compromise? Cambridge Uni
Press, C ambridge.
Peterson, A. Navarro-Sigiienza & H.
Diaz. 1998. The — for continued scientifie collecting:
\ geographic analysis of Mexican bird specimens. Ibis
140: 288—292.
Peterson, K. J. & E. H. — 2000. Regulatory évo-
lution and the origin of the bilaterians. Proc. Natl.
Acad. E U.S.A. 97: — 33.
Pimentel. D.. L. Lach. R. Zuniga & 2000.
iim: ntal and economic costs p nonindigenous
the United States. BioScience 50: 33-05
Pollock. "р, Э ЧЕ: . P. Cum-
mings. Sol A case for evolutionary genomics and the
со ало е of sequence diversity. Mo-
iol. vol. : 1776-178 Ө.
Prance, G. т E
— M velopment.
490—
Benitez-
. Morrison.
—
Eisen, N. А.
э. Syste matics, conservation and sus-
Biodiversity & Conservation 4:
$^ Beentje, J. Dransfield & R. Johns. 2000. The
tropic al es remains undercollected. Ann. Missouri
Bot. Gard. 87: rae
Raff, R. А. 1006 tthe Shape of Life. Univ. Chicago Press.
thicago.
Ganzhorn.
—
*
— . M. Goodman & J. U.
2000. Taxonomic revision of mouse lemurs ( os Me
in the western portions of Madagascar. Int. J. Primatol.
21: 963-1019
Richardson. P. M. 1996.
тары Introduction. Ann. Missouri Bot.
l-
Agenda 2000
Gard. 83
The Systematics
2,
2000. Our unknown planet: Recent discoveries
and the future. Introduction. Ann. Missouri Bot. Gard.
87: 1-2.
Rosen, D. E. 1978. Vicariant patterns and prb “al ex-
planation in — Syst. Zool. 27: 159-188.
Rossman, A. Y. & D. R. Miller. 1996. — matics solves
— ms in ape ulture wal forestry. Ann. Missouri Bot.
Gar d. х 17
Ruedas A.. T — zar-Bravo, J. W. Dragoo & T.
Yales. od The importance of being earnest: W e
anything, constitutes a de
à "specimen examine ad?”
29-132.
Ryder, O. A. 1986. Species conservation and systematics:
. Wheeler. 1990. An —— = -
The dilemma of subspecies. Trends Ecol. Evol.
10.
Savage, J. M.
)5. pou and the biodiversity cri-
sis. BioScience 45: 673-679.
Simpson, B. B. & J. i 1995. Systematics: The sci-
ence of MICE BioScience 45: 670-072.
Soberón, J.. J. B. Llorente & L. Onate. 2000. The use of
specimen- — databases for conservation purposes: An
example using Mexican M and pierid basher
flies. Hives & 41
E. 1999, The D T global biodive rsilv
3-32 í
Conservation 9: 1441—1466.
and its ie E Pp. Cracraft & F. T. Grifo
. The Living Blasi in € risis: Biodiversity Sci-
. Press. New Yor
Systematics Age nda 2000. 1994a. Systematics Agenda
2000: Charting ш Biosphere. Systematics Agenda
2000. Society of Systematic Biologists. American Soci-
ety of Plant к a
ciation of Systematics р. tions, New Yor
1994b. Systematics Agenda 2000: Charting the
Biosphere. Technical Report Systematics Agenda
2000. Society of Systematic Plucis, American Soci-
ety of Plant Taxonomists, Willi Hennig Society, Asso-
ciation of Systematics C s sctions, New York. | Reprint-
Willi Hennig Society, Asso-
К.
ed. with corrections, 5.]
Systématique Agenda 2000. 1994.
sphere. Société Frangaise de
8 А2000 — report, French клк
Thalmann, U. T. Geissmann. 2000. КИ and
variation in the western woolly lemur (Avahi
species — uni-
— levé de la Bio-
sté са Paris
geographic
decida ntalis) with en of a new
color). Int. J. ges 21: 915-941.
Vane-Wright. R. I. 1996.
Systematic s and the conse وا
of biological diversity. 47
Ann. Missouri Bot. Gard. §
51.
Vecchione. M. & B. B. Collette. 1996. Fisheries agencies
m marine biodiversity. Ann. Missouri Bot. Gard. 83:
1994.
Conservation Biol. 8: 351—
Diagnosing че of
›.
— A. P. «€ R. DeSalle.
consery ation management.
W he e dl т, Q. D. 1905a. Systematics, the scientific basis for
inventories of biodiversity. Biodiversity & Conservation
1: 176— 189.
. 1995b. Systematics and biodiversity:
BioScience Su uppl. 19 —
of taxonomy in gene etic resource
Pp. 59-70 in K. E. Hoagland & A. Y
Rossman (editors), Global Genetic Resour
Ownership, and E re Property Rights. Associa-
Washington,
Policies at
higher leve ‘ls. Ө.
9‹
management.
Acce ә.
—
—
tion x Syste matics Collec
& J. Cracraft. 1996. amó preparedness: Are
we — to meet the biodiversity — Pp. 435-
H6 in E. О. Wilson. M. Kudla-Reaka &
(editors). Biodiversity H: Understanding and ba ting
our a al Resources. National Academy of Scienc-
ES
Wilson
Press. Washington. rar
n E Meier. 2000. Species Concepts and Phylo-
genetic Theory: A ا Columbia Univ. Press. New
Yor
& N. I. Platnick. 2000. The phylogenetic spec les
concept (sensu Wheeler and Platnic ae 55-09 in Q.
D. Wheeler & R. Meier (editors). s Concepts and
dii snetic Theory: A Debate. Сайыш Univ. Press,
w York
“у, R. A. 1999. Species: New Interdisciplinary Es-
savs. MIT Press. Boston
144
Annals of the
Missouri Botanical Garden
Wolfenbarger, L. L. & P. R. Phifer. 2000. The ecological
risks and benefits of + tically engineered plants. Sci-
ence 290: 2088-206
Internet Resources
[All Species Inventory |: http://www.all-species.org/
[DIVERSITAS]: http://www.icsu.org/diversitas/
|G BI К]: htty // rg/
[GenBank]: htt ыызы: ne ebi. nlm.nih.gov/
[С i Global Taxonomy Initiative: http://www. biodiv.org/
E ht ml
»ec-tax/gti/
[ Dus BASE |: ip ihren harvard.edu/treebase
[ pis of Life] NSF _ Workshop —
esearch.amnh 1 nl
O
а
TAXONOMY AND HERBARIA
IN SERVICE OF PLANT
CONSERVATION: LESSONS
FROM MADAGASCAR’S
ENDEMIC FAMILIES!
George E. Schatz?
ABSTRACT
Our actions during the next two decades will largely determine how many of the world’s ca.
fundamental data that define both the t
vascular plants will survive for future generations. The
within which species are cireumseribe d and delimited ч related 5
These herbaria, and the taxonomists who
of those species, reside in the € s ca. 3000 herbaria.
and must play a critical role i
project that has focused on the species in М
of reviewing, and then revising when necessary, the exist
referencing of primary occurrence data га facilitated simple
entifying as rapidly as possible those species
300.000 species of
taxonomic frameworks
the geographical —
work 1 the m, can
)ecies, as well as
GIS analvses of Exter
labs ence in ш.
estimates of the number of * ‘subpopulations,
species in the endemic families re
tance: both those already incorporated in the protected areas system,
Partnerships among the world’s herbaria can ef
outside of the protection network.
— and thei AUT presence
assessment of extine tion risk. In s NA perse ne mapping of Pall
assessment of the most threatened vascular plant species by focussing on taxa endemic at politic ‘al and regional (e.g.,
Hotspot) scales. The synthesis and analysis of the primary data house ai in the world’s
record of plant life on
planning. and thereby minimizing the loss of plant diversi
ey words:
Earth—constitutes the most effec
herbaria—our only incontestab da
tive and robust means of directly informing conservation
conservation, natural history collections, primary occurrence data, taxonomy.
As the term “biodiversity” has entered the in-
ternational lexicon of the environment during the
past decade, one of the most fundamental aspects
of the concept has often been neglected: principal
knowledge of biological diversity emanates from the
study of natural history collections by taxonomists.
Plant taxonomists utilize herbarium specimens to
егесі taxonomic frameworks within which species
are defined, circumscribed, and delimited from re-
—
ated species. In addition to portraying the limits
of variability of individual taxa, those herbarium
specimens representative of a given taxon also con-
stitute primary occurrence data that collectively de-
pict the verifiable (i.e... vouchered) geographical
and the
the
distribution of the taxon. Thus, herbaria,
taxonomists who examine, classify, and curate
primary specimen data that reside in he rbaria. fur-
"what," “where,” and "when." initially
necessary to document plant life on Earth. As such.
nish the *
they have a unique and critical role to play in glob-
al efforts to mitigate the loss of biodiversity. Recent
studies of the families of vascular plants endemic
2000a) demonstrate
that central to such a role is the revisiting, reattri-
to Madagascar (Schatz et al.,
buling, and synthesis of the primary data itself.
Tens of thousands of plant species have already
been listed as threatened with extinction. The 7997
IUCN Red List of Threatened Plants (Walter & Gil-
let, 1998). The World List of Threatened Trees (Old-
field et al.. 1998), and their to-be-revised-vearly
electronic Internet successor, 2000 [UCN Red List
of Threatened Species (see [IUCN]).
—
serve as sober-
с Birkinshsw, F.
Magil
'T thank N Andrianjafy, (
vrl in the k ya mic F Ke Pojat R.
IL and an anonymous reviewer
> Missouri - Batani cal Gar
t Piscic
КОНЕ
tesies extended by the Government of — ar (Direct tion Générale de
Protégée
mal Science пие grants DEB-9024749 and DEB- 062707 2 and from the John D. and Catherine
z Claiborne and Art Ortenberg — LWO Inc
. Louis. Missouri 63166-0299, U.S.A. george.se hats @шороп: Git
the Association Nationale pour la Gestion des Aires 1
Nati
Foundation, the
2 Missouri — “al Garden, P.O.
Box 299,
ANN. Missouni Bor.
Randriantafika,
| for generating е
г for helpful comments on the — ript. F y vork was
and the Pare Botanique et Zoologique de Tsimbazaza and the
аот ану for 6x ir partic-
^ P. Lowry
— and Y
S statistics: and V. Hollowell.
bo бг — eder
Antananarivo, Madagascar. We gratefully acknowledge cour-
a Gestion des Ressources Forestiéres) and by
This research was conducted with — from U.S.
. Mac Arthur
. and the National Geographic Society
GARD. 89: 145-152. 2002.
146
Annals of the
Missouri Botanical Garden
Table 1
Malagasy endemic plant families: Revised (since 1998) taxonomic frameworks.
No. of Newly ewly Placed into
species recognized described synonymy
Asteropeiaceae 8 3 (38%) 2 (25%) —
Didymelaceae 2 —
Kaliphoraceae l = —
Melanophyllaceae 6 | (17%) 1 (17%) 4
Physenaceae 2 —
Sarcolaenaceae 60 25 (42%) 23 (38%) —
Sphaerosepalaceae 18 2 (11%) — —
Tota 97 31 (32%) 20 (27%) 4
ing baselines on the state of the Earth’s plant re-
sources. As commendable as these efforts are, the
fact remains that the majority of plant species have
not yet been evaluated with regard to their risk of
extinction. As would be expected, the lists are
heavily biased in their representation to the com-
paratively well-known floras of Europe and North
America, and to a lesser degree Australia and
South Africa, or to specific taxonomic groups (e.g..
conifers, palms) or life forms (e.g., succulents) pop-
ular in cultivation. One suspects therefore that the
number of threatened plant species identified thus
far is only the tip of the iceberg. In the absence of
an exhaustive survey of all ca. 300,000 (or more?
species of vascular plants (Prance et al., 2000), a
strategy to identify and safeguard those remaining
species most at risk of extinction must be imple-
mented as rapidly as possible. How can the daily
work of herbaria and taxonomists contribute to such
a strategy and thereby directly inform conservation
—
planning?
CONSERVATION ASSESSMENTS BASED ON OLD
TAXONOMY WILL BE INADEQUATE AND
MISLEADING, BECAUSE THEY FAIL TO INCLUDE
RECENT PRIMARY DATA: REVISITING THE “WHAT”
The goal of the Endemic Families of Madagascar
Project has been to provide current assessments of
the conservation status of all species in the seven
families of vascular plants endemic to Madagascar
and the Comoro Islands (Schatz et al., 2000a). An
essential first step in that process was the reeval-
uation of existing taxonomic frameworks for each of
the families, four of which were originally sub-
sumed within non-endemic families: Asteropei-
aceae within Theaceae (Perrier de la Báthie, 1951):
Kaliphoraceae Melanophyllaceae within
Cornaceae (Keraudren, 1958);
within Capparaceae or Flacourtiaceae (Perrier de
a Bâthie, 1946). The available taxonomy prior to
our review dated on average to 1952 (ranging from
and
and Physenaceae
—
1937 to 1963). Reevaluation of existing taxonomic
frameworks entailed examination of all primary
specimen data. In the case of three of the fami-
lies—Didymelaceae (Leandri, 1937), Kaliphora-
ceae, and Physenaceae—together comprising just
five species, the most recent frameworks remain
valid. But in the remaining four families—Astero-
peiaceae, Melanophyllaceae, Sarcolaenaceae (Ca-
vaco, 1952). and Sphaerosepalaceae (Capuron,
1963)—review of all primary data necessitated
their taxonomic revision (Randrianasolo & Miller,
1999; Schatz et al., 1998, 1999a, b, 2000b, 2001:
Lowry et al., 1999, 2000). In all but one instance,
a synoptic format for the revisions was adopted for
the sake o the taxonomic
framework was illuminated through new identifi-
cation keys emphasizing diagnostic features, and
full synonymy for each taxon was given, but only
the newly described species were provided with full
expediency: revised
descriptions. These revised taxonomic frameworks
include 31 newly recognized species (or 34% of the
92 total species now enumerated in those four fam-
ilies), 26 of which were newly described; 4 formerly
recognized species and a number of infraspecies
were placed into synonymy (Table 1). Primary data
that had been collected since publication of the
former frameworks accounted for 19 (73%) of the
26 newly described species. In the remaining 12
newly described or recognized species, specimens
ascribed to other species in the previous taxonomy
were judged to represent clearly discernible, dis-
tinct taxa based upon morphological and ecogeo-
graphic criteria. Conversely, in the cases of species
and infraspecies now placed into synonymy, for-
merly cited specimens, in conjunction with newly
available material, revealed taxa that could no lon-
ger be distinguished from earlier described species.
Clearly, assessments of conservation status based
on the taxonomies that existed prior to our reeval-
uation would have been highly flawed and mislead-
ing, and would have failed entirely to account for
Volume 89, Number 2
2002
chatz 147
Plant Conservation
one third of the species now recognized in the sev-
en endemic families.
CONSERVATION IS ALL ABOUT GEOGRAPHY, BUT
Most OF PLANT DIVERSITY 15 Nor GEO-
REFERENCED: REATTRIBUTING THE “WHERE”
Mapping the primary occurrence data upon
which a taxonomic framework stands has long been
a preoccupation of taxonomists: most revisions in-
clude distribution maps of the taxa. Distribution
mapping is usually a tedious task insofar as the
majority of primary data collected prior to the latter
decades of the 20th century are only imprecisely
located in relation to population centers or phys-
iographie features of the landscape, but lack the
precise geographical coordinates that can now be
assigned so easily in the field with Global Position-
ing System (GPS) technology. As a consequence,
most of plant diversity must be post facto geo-ref-
erenced in order to capitalize on the current revo-
lution in computer mapping and spatial analysis
brought about through Geographical Information
Systems (GIS
data representing the Malagasy endemic families
tools. Geo-referencing of
~
primary
has been greatly facilitated by the compilation of a
Gazetteer to Malagasy Botanical Collecting Locali-
ties [Schatz & Lescot, 2001]. Some historical spec-
imens that lack coordinates may simply be impos-
sible to geo-reference to a seale that has meaning
for conservation. In the majority of cases, however,
geo-referencing can be achieved to obtain a Mini-
mum Mapping Unit of plus or minus one minute of
latitude/longitude, more than sufficient for the
coarse area estimates used in conservation assess-
ments. Even in cases of uncertainty involving col-
lections known to originate from within protected
areas in Madagascar, but without any greater pre-
cision, e.g., the Réserves Naturelles or RN collection
series, mapping to the centroid of the protected
area polygon still constitutes a valuable reattribu-
tion that can be included in area and subpopulation
estimates. Of even greater importance is that col-
lections in the RN series document historical pres-
ence within the protected areas networ
With
works in place that are based on the totality of
new and reconfirmed taxonomie frame-
=
available primary data, the next step in assessing
the conservation status of each species entails anal-
vsis of its geography, including the geography of
current land cover and threat. Conservation ulti-
mately depends on sufficient space to maintain a
viable population, the size of which will vary con-
siderably from one species to another depending
upon life histories and habitat requirements. Al-
though ideally one would want to conduct a long-
term Population Viability Analysis (PVA) to deter-
mine exactly what constitutes “sufficient space" for
each species, area measurements calculated di-
rectly with GIS tools based solely on the primary
occurrence data can provide a rapid first estimate
of a species’ vulnerability to threat, and therefore
a means of prioritizing concern. Among the as-
sumptions inherent in such an initial evaluation
are: (1) that the known collections and localities for
a given species are a valid reflection of its abun-
dance and distribution; (2) that widespread and
common species will be at lower risk than restricted
and rare species; and (3) that species occurring
within protected areas will be at lower risk than
species that occur only outside of protected areas.
The IUCN Red List Categories (IUCN, 1994; IUCN/
SSC Criteria Review Working Group, 1999) serve
as a guideline for incorporating the area measure-
ments of Extent of Occurrence and Area of Occu-
pancy into a hierarchical delineation of extinction
risk. For example, among the species in the endem-
ic Malagasy families, newly described Melanophyt-
la modestet G. E Schatz, Lowry & . Wolf (Me-
lanophyllaceae) (Schatz et al., 1999). is classified
as Critically Endangered (CH) given an Extent of
Occurrence less than 100 km’, an Area of Occu-
pancy less than 10 km?, and a single known pop-
ulation (Schatz et al., 2000a). On the other hand.
\steropeia densiflora Baker is classified as Vulner-
able (VU) due to an Extent of Occurrence less than
20.000 km? (Fig. 1), an Area of Occupancy less
than 2000 km? (Fig. 2), and less than 10 known
“subpopulations” (the number of non-contiguous
occupied cells or cell clusters within the 10 km X
10 km grid utilized to estimate the Area of Occu-
рапсу) (Fig. 2) (Schatz et al., 2000a).
In addition to facilitating rapid area measure-
ments that can be incorporated into evaluations of
extinction risk, reattributing primary data with geo-
graphical coordinates also allows the modeling of
potential distribution (Skov, 2000; see also [BIODI]
and [Species Analyst]. Utilizing the spatial analyt-
ical functions of GIS, an envelope of independent
physical and environmental variables associated
with a set of primary occurrence data points can be
described, and in so doing define an environmental
niche, and hence, a potential distribution that cor-
responds to the range of documented heterogeneity.
When compared against current land cover as re-
vealed from satellite imagery, remaining viable
habitat coinciding with (preferably recently) docu-
mented distribution can be identified. In the ab-
sence of extant viable habitat that intersects the
documented distribution. 1.e.. when all historical
148
Annals of the
Missouri Botanical Garden
Figure ent of Occurrence (= 17.830 km?) of
Asteropeia pie чөн, Baker in Madagascar.
=
primary data originate from areas now converted to
urban or agricultural uses, conservation hopes for
a species may still lie within areas of potential dis-
tribution and remaining viable habitat.
Conservation in such places as Madagascar may
ultimately depend upon encompassing a species
within some type of protected area. An overlay of
protected areas polygons in relation to primary oc-
"gap"
tection exists. Among the 97 species in the seven
currence data instantly reveals if a in pro-
families endemic to Madagascar, at least 28 species
are not presently recorded from protected areas.
The situation is potentially far bleaker for the spe-
75% of the
4011 endemic species are not yet documented in
protected areas (Valencia et al., 2000). Similarly,
the 1995 revised ROTAP (Rare or Threatened Aus-
tralian Plants) listing (Briggs & 1996) re-
veals that 47% of the 5031 listed taxa are not doc-
cies endemic to Ecuador, where over
Leigh,
=
umented from protected areas. If one
conservation is to ensure that the maximum number
of species is included within the protected areas
network,
dictate that centers of endemism,
then. considerations. of. complementarity
L.e., concentra-
tions of co-occurring local endemic species, be
identified (Williams, 1999).
multaneous mapping of all species in the endemic
For Madagascar, si-
families revealed, for example, that the Ibity/Itremo
goal o
Figure 2.
sis of * al y eer (7 individual cells + 2 cell «
steropeia densiflora Baker in — ar as
10 km grid.
Area of Occupancy (= 1800 km? ) € num-
‘lus-
ers = 9) of /
de fined by a 10 km X
Massifs on the Central High Plateau, which are
geologically characterized by a complex mosaic of
granitic, marble, and quartzite substrates, consti-
tute an important center of endemism that is not
currently encompassed within the protected areas
network. Of the ten endemic families’ species re-
corded Ibity/Itremo, five (all Sarcolaena-
ceae)—Leptolaena diospyroidea (Baill.)
Pentachlaena latifolia H. Perrier, Perrierodendron
— еи -F. Leroy, Lowry, Haevermans, Labat
& . Schatz, Schizolaena microphylla H. Perrier,
from
Cavaco,
anda an quede species of Xerochlamys/Sarco-
are essentially restricted to the Massifs,
laena
and are therefore entirely lacking protection. The
presence of numerous other local endemics in the
Ibity/Itremo Massifs,
gumes (Du Puy & Moat,
including a number of le-
1998), forcefully argues
for the immediate establishment of new protected
areas in the region. Simultaneous mapping of all
species in the endemic families also revealed the
importance and management needs of existing pro-
tected areas. In particular, the small Réserve Na-
turelle Integrale of Betampona (2228 ha) shelters
20 species in the endemic families, including two
newly described as a result of revised taxonomies
within Sarcolaenaceae (Pentachlaena betamponen-
sis Lowry, E. Schatz and
Rhodolaena leroyana С. E. Schatz, Lowry & А.-К.
Wolf) that are known only from this reserve. Map-
Haevermans, Labat & G.
Volume 89, Number 2
2002
Schatz
Plant Conservation
149
ping reattributed primary occurrence data of re-
stricted range species, i.e., country endemics. о
species endemic to. Ecoregions (Olson & Dinner-
stein, 2002 [this volume]) or Hotspots (Mvers et al..
2000).
robust. means of directly informing conservation
may well constitute the most efficient and
planning.
Much or PLANT DIVERSITY IS “RARE.” LE.
DOCUMENTED BY VERY FEW COLLECTIONS, AND
HERBARIA AND THEIR ASSOCIATED DATA ARE
IDIOSYNCRATIC: STRATEGIES AND NEW
STRUCTURES FOR RAPIDLY ACHIEVING THE
SYNTHESIS OF REATTRIBUTED PRIMARY DATA
Faced with the very real prospect of losing a sig-
300.000 or
more species of vascular plants during the coming
nificant proportion of the estimated
decades. it is incumbent upon the systematics com-
munity lo synthesize the most relevant primary
data. and disseminate that data to governments and
the conservation community. By definition, such a
synthesis must go beyond Red Lists per se as they
are currently envisioned by IUCN (see [IUCN |). Al-
though Red Lists attempt to draw attention to the
species most at risk of extinction, because they lack
the underlying geo-referenced primary data. thes
have only very limited utility for conservation plan-
ning. Nevertheless, Red Lists in their current form.
along with country and. regional checklists. can
help guide the prioritization of primary data syn-
25% of Ecu-
adoran species known only from the type collection,
thesis. By identifying the greater than
and the greater than 50% known from only one or
two populations, and conversely, the less than 10%
that are “common” (known from ten or more pop-
ulations), the Ecuador Red List of Endemic Plants
Valencia et al.. 2000) serves to prioritize which
species should be subject to comprehensive. pri-
ROTAP lists nearly
200 Australian species known only from the type
1990).
prehensive synthesis of all vascular plant species
mary data synthesis. Similarly.
collection (Briggs & Leigh. Surely, a com-
known onlv from their type collection. and the geo-
referencing of as many as possible, must be one of
the very first priorities.
Species known only from a single (tvpe) collec-
tion are merely an extreme case of endemism.
Within the context of the Convention on Biological
Diversity (see [CBD]). governments are obligated to
pay particular attention to all those species endem-
іс within their borders. Among the nearly 54.000
species listed in the 1997 IUCN Red List of Threat-
ened Plants (Walter & Gillet, 1998), 91% were sin-
ele-country endemies. The recent Red Book of Iran
(Jalili & Jamzad, 1999) evaluates all 1727 endemic
species (22% of the total flora). and the Ecuador
Red Book focuses solely on the endemic species.
What is required next is the assimilation and geo-
referencing of the primary data of all of these. as
well as every other country’s endemic taxa. Priority
should be accorded to species representative of en-
demic higher taxa, i.e., families and genera. A sim-
ilar approach should be adopted within the frame-
work of the new Critical Ecosystem Partnership
Fund initiative (see [CEPF]),
dress conservation needs at a regional scale, as de-
which seeks to ad-
limited by conservation Hotspots. which are them-
selves defined in part by a minimum number (1500)
2000).
Knowing where the endemic species occur (or al
of endemic plant species (Myers et al.,
least occurred at some point in the past, and might
possibly still occur) within countries or regions 15
fundamental for the rational allocation of finite con-
servation resources
But as the review of all primary data represent-
ing the Malagasy endemic families has demonstrat-
ed. knowing where the endemic species occur must
he predicated on knowing just what the endemic
species are. Catalogues such as those for Ecuador
1999) (Brako
& Zaruechi. 1993) involve some review of the pri-
(Jorgensen & León-Yánez. and Peru
mary data. but generally do not exhaustively inven-
tory all existing specimens. Revised taxonomic
frameworks, however, should in theory be based
upon examination of all existing collections. and
therefore represent the most appropriate and op-
portune point at which to disseminate reattributed
primary data. Indeed, deposition of reattributed pri-
mary data into an Internet-accessible database
should be a sine qua non for publication of a re-
vised taxonomic framework, just as deposition of
GenBank
has become (in most cases) a
nucleotide data into (see
(NCBI)
precondition for publishing
sequence
necessary
phylogenetic frame-
works. For revised taxonomies of the Malagasy en-
demic. families. comprehensive, reattributed (i.e.
geo-referenced) primary data have been deposited
in the world’s largest botanical specimen database.
to which Internet access is provided through
W"TROPICOS (see [MBG]). Recent discussion of
the state of bioinformatics for biodiversity has
sounded the call for improved infrastructure, and
highlighted various developments involving remote
query and retrieval from multiple, so-called distrib-
uted databases (= “interoperability”) (Bisby, 2000:
2000: Krishtalka & Humphrey.
Nevertheless. in conjunction. specifically
Edwards et al..
2000).
with the publication of revised taxonomic frame-
works, it would seem appropriate and extremely
150
Annals of the
Missouri Botanical Garden
Table
2. Representativeness of TROPICOS specimen
database on 10 October 2000.
1.5 + million specimen records
e 922,068 geo-referenced
e |. a million identified to spp. representing 117
SF
.806
752 spp. (39%) represented by only р specimen
ө ub 709 spp. (53%) represented by | or 2 specimens
€ 96,456 spp. (82%) represented by 10 or less specimens
@ 4,742 spp. (4%) represented by 50 or more specimens
useful lo establish a GenBank analog
(“SpecimenBank”) for deposition of the underlying
geo-referenced primary data. The utility of recent
Species Plantarum—Flora of the World (see
[ABRS]) treatments of Irvingiaceae (Harris, 1999),
Stangeriaceae (Steyn et al., 1999), and Welwitschi-
aceae (Steyn & Smith, 1999) would be significantly
enhanced if the underlying primary data were ac-
cessible from an Internet specimen database. Just
as the nucleotide sequences in GenBank document
the microgeography of biodiversity, natural history
collections define its macrogeography. Obvious
linkages between the two scales should be made,
as well as to seedbanks and living germplasm col-
lections (at [IPGRI] and [NPGS].
Although the TROPICOS database of the Mis-
souri Botanical Garden does contain some exhaus-
tive sets of primary data such as those assimilated
in the course of the review and revision of the Mal-
agasy endemic families, in general, it reflects the
idiosyncratic, incomplete nature of herbaria and
their associated specimen data. The sources of the
representativeness of TROPICOS (mostly contem-
porary collections from a limited number of re-
gions), wherein 117,806 species are represented by
1.5+ million specimens, dictate that the majority
of species are represented by only one or two spec-
imens; conversely, only 4% are represented by 50
or more specimens (Table 2). Just as taxonomic re-
visions require the pooling of primary data from
numerous herbaria, the synthesis of such data for
conservation must also involve “North/South” her-
barium databasing partnerships. For given
country or region of the world, usually a limited
any
number of internal and external herbaria hold the
majority of unique primary data. Adopting a
cies by species approach, and beginning with those
spe-
endemic to countries and regions, North/South her-
barium partnerships should work to synthesize pri-
mary data into a global plant conservation database.
A model already exists to emulate for the devel-
opment of such a database. Botanical Garden Con-
servation International (see [BGCI]) serves as a co-
ordinating body to galvanize ex situ plant
conservation efforts at over 500 member botanical
gardens, maintaining a database of the ca. 85,000
species currently in cultivation in those gardens.
Similarly, the 3000 herbaria worldwide (Holmgren
et al., 1990), and indeed, the ca. 215,000+ plant
species not yet in cultivation, would benefit tre-
mendously from an analogous coordinating body
“Herbarium Conservation International”) to help
facilitate databasing partnerships for the synthesis
of primary data that could directly inform in situ
plant conservation. Such a coordinating body would
assist in the organization of regional herbarium net-
works such as the highly successful Southern Af-
rican Botanical Diversity Network (see [SABO-
NET]. and ensure that they are partnered with the
appropriate Northern herbaria. The ongoing Red
ist Program within SABONET (Golding. 1999a,
1999b, 2000) would be greatly enhanced from for-
malized partnerships with the Northern herbaria
where the majority of primary data from the region
are housed.
The task of synthesizing primary data of the most
threatened plant species is large, but by no means
insurmountable. The analysis of over 4000
adoran endemic species was achieved within a little
over a year of the publication of the Ecuador Cat-
alogue. With ca. 300,000 species to track, and
3000 herbaria, each herbarium would need to take
responsibility for only 100 species on average. /n-
dex Herbariorum lists 8800 staff working at the
3000 herbaria: throw in an additional 1200 stu-
dents and volunteers, and each person would be
responsible for collating the primary data of just 30
species on average. In fact, the task is not even
that great. There are numerous widespread species
of little or no conservation concern (except when
their invasive capacity leads to the displacement of
indigenous species). The problem should thus be
attacked from both ends, identifving both the most
widespread and "weedy" species, as well as those
represented only by one or several collections. Mo-
mentum is building to synthesize information on
invasive species globally (see [GISP] and [NBII]).
but there is as yet no organized effort to tackle the
atter, i.e., the rarest of species. Similarly, great pro-
gress (sometimes even with redundant and —
ping efforts) is being made to diffuse the “1
of biodiversity (see [ABI], ПРА, "ПОР, TITIS),
[Species 2000)).
servation organizations seek to prioritize and pro-
). However, as governments and con-
tect remaining tracts of viable habitat, it is imper-
ative that they have access to more than just the
have
“what” of threatened biodiversity. “Names”
meaning only in relation to the primary data that
Volume 89, Number 2
2002
Schatz
Plant Conservation
define them: originally by their types, and then sub-
sequently all the other specimens assigned to them
by taxonomists in the course of trying to order the
grand diversity of life. Therefore, by definition, the
meanings of plant “names” cannot remain static.
Each new collection expands the meaning in space
and/or time, and has the potential to significantly
modify or alter the meaning, or even to define a
new name. To maximize the number of plant spe-
cies that will survive the current extinction wave,
m 9
we must also furnish the “where” of rare and threat-
ened plant species, continually revisiting and
attributing the primary specimen data, our only in-
contestable record of life on Earth.
Literature Cited
Bisby, К. А. 2000. The quiet pn Biodiversity in-
formaties and the Internet. Science 289: 2309-2312
Brako, L. & J. Zarucchi cand 1993. Catalogue of
the Flowering Plants and Gymnosperms of Peru. Mon-
3
ogr. Syst. | Missouri Bot. Gard. 45: i-xl, 1-1280.
Briggs. J. D. & J. I eigh. 1996. Rare or Threatened
—
stralian Plants, 1995 rev. ed. CSIRO, Australia.
Capuron, R. 1963. Rhopalocarpacées. Fl. Madagasc. 127:
1-42.
Cavaco, A. en wes Fl.
Du Puy, D. J. & J. Moat. 199€
classification. in
Madagase. 126: |
. Vege — mapping abd
Madagsas ar ene GIS
mplications
and recomme nations for the conse — of biodiver-
v. Pp. 97-117 in € uxley, J. M. Lock &
Culler — Chorology, Taxonomy, and Ecology af
the Floras of Africa and Madagascar. Royal Botanic
, Кем
uw
. Lane & E. S. Nielsen. 2000. Inter-
ope кү at biodiversio databases: Biodiversity infor-
on every desktop. Science 289: 2312-2314.
9‹ А — Red Data List for southern
\frica SORORE News 4: 111-118
Е. 999}. Не pus and Red. Data Lists. SABONET
Mews |: 202-20
2000. De aa with DDs: i neertainties m E Б
ficiencies in — Listing. SABONET News 5: 28—
Harris, D. J. ‚ Irvingiaceae. Species LUE я
tralian — Resources Study, Canberra
Holmgren, P. K.. N. Holmgren & . Barnett (edi-
tors). 1990. Index Herbariorum: Part t The Herbaria
of the World, 8th ed. International Association for Plant
mation
Taxonomy and New York Botanical Garden. Bronx. New
York.
IUCN, 1994. IUCN Red List Categories. IUCN. Gland.
Switzerland.
IUCI N/SSC Criteria Review Working Group. 1999. IUCN
Red List criteria review provisional report: Draft of the
— changes and recommendations. Species 31—
32:
g A ЧЛ Jamzad. 1999, Red Data Book of Iran. Re-
rch perce of Forests and Rangelands,
мя nsen, P. M. & S. León-Yánez (editors).
ogue of the Vascular Plants of к
Bot. Missouri m Gard. 75: i—vii, 1—11
ke — n, М. 1958. Cornacées. Fl. —
Tehran.
1999, Cata-
аа Syst.
158: 13-
— . & PS. Humphrey. 2000. Can natural his-
tory museums capture the future? Bioscience 50: Oll-
617.
Leandri, J. 1937. Sur l'aire et la position systé e du
3
genre Баре s — Thouars. Ann. Sei. Nat. Bot.,
sér. 10, 19: —319.
Lowry. P. P., Il. n J.-N. Labat, С. E. Schatz,
E Leroy & A.-E. Wolf.
Madagascar. V. A s Pic
2000. families of
revision of Kremolaena. Pen-
tachlaena and ʻ ian (Sarcolaenaceae). Ad-
ansonia, sér. : 11-31.
‚ G. К. Se а J.-F. Leroy & A.E.
Endesa families of Madagascar. Ill.
sion of Schizolaena (Sarcolaenaceae).
21: 183-212.
Mvers, N., R. var ler, ne
it. 2000
da Fonseca & J. Ken
—
Wolf. 1999.
А synoptic revi-
Adansonia, sér. 3.
м 2
. Mittermeier,
ES sity Кыр bor
: 853—858.
. Mac — mn. 1998.
е vis — 40:
Oldfie lk м The World
. Lusty
List « рез ne a Tre es. World Conservation Pres
( ч жн U.K
Olson, D. M. & E. Dinerstein. 2002. The Global 200:
Priority ecoregions for global conservation. Ann. Mis-
souri Bot. Gard. € “ E. 9-22
Perrier p : Báthie. Н. 1946. Flacourtiacées. Fl. Mada-
gasc. : 1-13
T 23 Thea ‘асве Madagasc.
13.
Prance, G. E, H. Beenie, i — i — 2000.
The tropical е remains undercollected. Ann. Missou-
T"
ri Bot. Gard. 87: 67-7
Randrianasolo, А. 8 J. S. Miller. 1999. ‘Taxonomic revi-
sion of the genus Sarcolaena (Sarcolaenaceae). Ann.
ТИЕ Вог. Сага. 86: 7
22.
— G. : ~ P. P. Lov 7. Wolf. 1998. Endemic
Madagascar. 1. — revision of Me-
amphi В r (Me Mor dic eae
. Adansonia, sé
—
999a. Endemie
\ — revision of Sphaerosepala-
Adansonia, sér. 3, 21: 10712
————. 1999b.
A
— IV. A syn
y od families of
Madagascar. П.
беае.
Ende mic families of
optic revision of . = (Аз-
3
y
teropeiaceae). Adansonia, sér. 3, ed 208.
. Birkinshaw, P. P. Lowry Il. ean i
& F. | TUN 2000a. The endemic jdm families of
Madagascar apes oq taxonomy and conser-
vation. Pp. 11-24 i . Loure | S. Goodman
(editors). Diversis T Rade mism in n Na ar. Mém.
Soc. Вторбо Er.
II & A.-E. Wolf. 2000b.
байа "s of Madagascar. VI. A synoptic revision of
22.
Endemic
Rho-
930
Adansonia, sér. 3, 22: 239—
. P. Lowry
dolaena (Sarcolaenaceae).
2524
. P. P. Lowry H & A.-E. Wolf. 2001.
ladagascar. УП.
Endemic fam-
ilies af 3 — revision of Lepto-
laena sensu ids ricto (Sare — naceae). Adansonia, sér. З.
23: 171-1
Skov, E
on climatic similarity. ' axon я 2 515.
A. Smith. . Welwitschiaceae.
з жа ub. al Resources
2000. Pote ntial plant distributia mapping based
ji )
Steyn, E. M. é
Plantarum.
Study. Canberra.
Spe cies
F. Smith & K. D. Hill. 1999. Stangeriaceae.
Spe cies жаа. Australian Biological Resources
— Canberra.
Valen ., N. Pitman, S. León-Yánez & P. M. Jorgensen
(e — 2000. 1. ‘bro rojo de las plantas e 'micas del
152
Annals of the
Missouri Botanical Garden
Ecuador 2000 [Red Book of Ecuador's Endemic Plants].
PUCE editions, Quito.
Walter, K. S. & Gillet (editors), 1998. 1997 IUCN
Red List of Threate ned Plants. Compiled by the World
Conservation Monitoring Centre. IUCN—The World
Conservation Union, Gland, Switzerland, and Cam-
bridge, U.
Williams; P. H. Area-
. Key sites for conservation:
selection me — biodive sity. Pp. 211-249 in G.
Mi Balmford & J. R. Ginsberg, Conservation
байл Univ. P Сап
асе,
ina MERE World.
bri ridg
ress,
Internet Resources
[ABRS]. Plantarum. Flora of the
WWW anbg gov | itm
Assoc GALÓN for Biodiversity Information:
ane
Species
a -
Garden С Somésvadion International.
I. org.uk
[BIODI]. BIODI. Advanced computational approaches
to environme Pia and biodiversity information.
(biodi.sds«
[С ed Каш о оп Biological Diversity. (www.biodiv.
[С E te critical Ecosystem Partnership Fund. (www.cepf.
ne s/he гаа htm)
[GISP]. Global а (jasper.
stanford.edu/G
[IOPI]. The E xs Checklist Project. (iopi.csu.
edu.au/iopi/iopigpe l.htr
[IPGRI]. Intern: ШШ Plant Genetic Resources Institute.
www.ipgri.cgiar.or
[IPNI]. International Plant Names Index. (www.ipni.org/)
[ITIS]. Taxonomic Information System
(www.itis.usda.gov/plantproj/itis/index.html)
[IUCN]. International Union for the Conservation of Na-
ture. Species Survival Commission. IUCN Red List of
Threatened Species. (www.redlist.org/)
[MBG]. W3TROPICOS. (mobot.mobot.org/W3T/Search/
vast. html
Species Program.
| NBII]. Invasivespecies.gov: The Nation's Invasive Species
Information System (invasivespecies.gov/index.shtml)
[NCBI]. GenBank overview. (http://www.nebi.nlm.nih.gov/
Genbank/GenbankOverview.html)
[NPGS]. National Plant Germplasm System. (www.ars-grin.
gov/npgs
[SABONET]. Southern — Botanical Diversity Net-
work a кик —
[Schatz, G. 2001 | Gazetteer to Malaga
Botanical ( ‘alle ting Loc liis (www.mobot. org/MOBOT/
research/madagascar/gazetteet
[Species 2000]. (http://www. sp2000. org/)
[Species Analyst]. (www.speciesanalyst.ne
—
ar
CHICAGO WILDERNESS: A Debra K. Moskovits,? Carol J. Fialkowski,2
ү E DIAM Ad
CONSERVATION! 5
ABSTRACT
^
=
=
ре
=
In a of diverse and determined organizations launched a new initiative in the Chicago region. Our
vision? Chicago Wilderness: a thriving mosaic of natural areas. connected by greenways and wildlife corridors; embed-
ded in the nation’s third largest metropolis. In this vision, the region's human communities reclaim a cultural tradition
of protecting and restoring these globally significant natural communities that enrich our lives. Today. more than 100
organizations join forces to transform this vision into reali ty. A regional Biodiver rsity Recovery Plan—the result of three
years of assessment and planning by scientists, land managers, educators, and policy y strategists—sets priorities and
"
determines the lines of action for the coalition. This жайна! agenda stems from our vision and recovery goals for each
ecological community; it encourages targeted research initiatives that focus on characterizing our native biological
diversity and on analyzin g elements critical to its recovery, Ultimately, though, the long-term survival of our natural
wealth rests on the support from "the public. " hile the challe ‘поез to conservation educators and Communicators are
many, QC hicago Wilderness аши us lo work toge ‘ther in unde rstanding our — nces, Channeling our resources, and
ff
ey words: Biodiver sity Rec covery Plan, Chicago Wilderness, citizen science, coalition, conservation education,
forest preserves, fragmentation, metropolis. mosaic, natural communities, oak savannas, oak woodlands, prairies, public
outreach, remnants, restoration, urban.
CHICAGO WILDERNESS: THE NATURAL CAPITAL west (Table 1). These remnants shelter a number of
species that are rare worldwide and that are listed
Chicago. the third largest metropolis in the Unit- as threatened or endangered in the states of Illinois.
ed States, is anything but "untrammeled by man. Indiana. and Wisconsin. These species once lived
where man is a visitor who does not remain" in the in the thriving, expansive grasslands and wood-
words of the Wilderness Act (Section 2.e: [Wilder- lands that have now been turned into monocultures
ness Асі). which has shaped our modern concept of corn or soybean.
of wilderness. Nonetheless, "Chicago Wilderness At the turn of the century, Chicago was blessed
—
is the emerging force in our urban landscape. with far-sighted planners who created a system of
Launched in 1996 by a coalition of 34 diverse and forest preserves in and around the city. In 1900.
determined organizations, Chicago Wilderness joins architects Jens Jensen and Dwight H. Perkins led
forces to conserve the globally important natural he Municipal Science Club in Chicago. which
communities that survive in our metropolis. worked hard for many years to ensure that the best
The surprise to many is that the boundaries of remaining natural lands were set aside for the pub-
Chicago Wilderness capture a spectacular array of lie (see Sullivan. 1997: Chicago Region Biodiver-
rare ecosystems: tallgrass prairies, oak woodlands. silty Council, 1999: 19-21). Famous architect and
oak savannas, sedge meadows, marshes, bogs, fens. planner Daniel Burnham and fellow architect. Ed-
and other prairie wetlands. In fact, within Chicago ward Bennet also contributed extensively to the
Wilderness—which extends in a crescent around cause. The wilderness in Chicago now survives in
Lake Michigan from southeastern. Wisconsin, parcels that range in size from half-acre remnants
through six Illinois counties, into northwestern In- to. the 19.000-acre Midewin National Tallgrass
diana—survive some of the best remnants of the Prairie. Some of the natural remnants survive in old
prairie and oak-savanna communities of the Mid- settler cemeteries or along railway rights-of-way.
! We thank the members of Chicago Wilderness and of its teams, along with the citizen scientists in the region.
Collectively, they are the power behind this re — effort and the focus of this paper
3E vironmental and e conservation Programs, The Field Museum. 1400 S. Lake Shote Drive, Chicago, Illinois 60605-
2496, U.S.A. dmoskovil Idmuseum.org. « mu Idmuseum.org
'* Department of cari The Field Museum, 1400 S. Lake Shore Drive. С hicago, Hlinois 60605-2496, U.S.A. gmueller
@fie — um.org
Conservation Programs, Chicago Zoological Society, Brookfield Zoo, 3300 Golf Road, Brookfield. Hlinois 60513,
l 4 \. кешу ТТИ СА ldzoo.org
ANN. MISSOURI BOT. GARD. 89: 153—163. 2002.
154
Annals of the
Missouri Botanical Garden
Table 1
oak savannas, oak woodlands, prairie wetlands) with high
floristic quality.
Remaining natural areas in Illinois (prairies,
Region Original acreage % Surviving
Galena 371,000 0.00005
Chicago 1,998,000 0.04.
West Central 3,217,000 0.00006
Springfield 2,041,000 0.003
Source: The Illinois Natural Area Inventory, Illinois De-
partment of Conservation, 1976
Most survive in the preserves that total
200,000 acres (81,000 ha) of federal, state, county,
and municipal protected lands and waters. All par-
cels exist within the much larger matrix of public
and private, built and natural lands that support
the region's 8 million residents.
Our natural communities suffer from a number
of stresses, including habitat fragmentation, inter-
ruption of natural fire regimes, air and water pol-
lution, and native and exotic invasive species. The
long-term survival of these endangered plant and
animal assemblages depends on the proper ecolog-
ical management and aggressive restoration of de-
graded lands, even—or primarily—in the region's
forest preserves and conservation districts. It is this
need for coordinated and rapid action that now
brings together more than 130 institutions includ-
ing landowners, government agencies at local, state,
and federal levels, education and research institu-
tions, and conservation organizations to turn the vi-
sion of Chicago Wilderness into reality.
AND WHY
CHICAGO WILDERNESS: WHAT, WHERE,
Chicago Wilderness is both the name of the co-
alition of organizations committed to protecting, re-
storing, and managing the natural communities in
the Chicago region, and the natural communities
themselves, along with the plants and animals that
depend on them for survival. Chicago Wilderness
also is the people of the region, the human resi-
dents whose quality of life depends on a thriving
regional nature reserve.
Urban living often results in an almost complete
detachment of people from the land. An important
goal of Chicago Wilderness is to reconnect a land-
less urban population with the pulse of nature. The
name Chicago Wilderness is meant in part to em-
phasize the connections between wilderness in the
inaccessible
conventional sense—in remote and
places—and wilderness in the places where people
ive and work.
| 1 work
But why strive for a regionwide coalition instead
over
of working through more manageable, individual
collaborations? The answer lies in the realities of
the region: extreme fragmentation of the natural
communities. The high-quality native patches in
the region are tiny. Nothing short of a massive, re-
gionwide, coordinated effort will ensure the long-
term survival of these natural communities, through
the management of much larger lands that surround
and connect the remnants. In the vision of Chicago
we reconnect the 200,000 acres of
fragmented natural areas through greenways and
Wilderness,
wildlife corridors. And in this vision, the human
communities reclaim a cultural tradition of restor-
ing, protecting, and managing our natural commu-
nities.
CHICAGO WILDERNESS: THE ORGANIZATION
Chicago Wilderness is a loosely structured or-
ganization, with an unconventional organizational
chart (Fig. 1). In Chicago Wilderness we are bound
by our common goals and objectives and by our
collaborative projects. We are in essence a volun-
teer-driven effort. Four teams focus on the central
lines of action: science, land management, policy
and strategy, and education and communication.
The teams attract participation of many non-mem-
ber institutions, which adds to the scope and
strength of the coalition. Current and past chairs of
these four teams and staff from other member or-
ganizations form the nucleus of a coordinating
group that develops central strategies, maintains
momentum, and resolves day-to-day problems. A
steering committee of executives in the Chicago
Region Biodiversity Council oversees the initia-
tives, including approval of budgets and projects.
The Council encompasses the chief executives of
federal, state, and local
all member institutions:
government agencies; land-owning agencies; re-
search and education institutions; and conservation
organizations. These executive, or voting, members
elect the steering committee. A proposals commit-
tee solicits and reviews proposals for priority con-
servation projects and recommends funding; the
committee has one or two representatives from each
And the Recovery Plan Task
Force spearheaded the gigantic effort of developing,
compiling, and writing the first draft of the Biodi-
of the four teams.
versity Recovery Plan for the region (see below).
A few ingredients have been instrumental in al-
lowing Chicago Wilderness to succeed as a massive
coalition. We briefly highlight these below.
(1) A critical mass of people thoroughly—pas-
sionately—committed to seeing the work
succeed, and ready to put an inordinate
Volume 89, Number 2
2002
Moskovits et al. 155
Chicago Wilderness
Figure 1.
—
—
N
w
—
Chicago Wilderness Structure
Chicago Region Biodiversity Council
Executive Council
General Members
Proposals
Steering Committee | J—
Committee /
Science
agemen
Collaborative Projects
amount of time and effort into making it hap-
pen. For us, this core energy came from the
coordinating group, with its 12-15 members.
Early, visible, and popular successes. In Chi-
cago Wilderness, these early successes Came
from some of our public awareness materials.
The Atlas of Biodiversity (Sullivan, 1997)
the first publication of the Chicago Region
the
natural communities in the region, with their
Was
Biodiversity Council. It characterizes
rich assemblages of plants and animals. and
helps bring these natural areas to the atten-
tion of a very broad audience. most of whom
are unaware of the existence of such biolog-
The Atlas
was one of the first concrete examples of how
ical riches in our own backyard.
Chicago Wilderness could help individual
members become more successful and effec-
tive in their work. The whole region needed
a powerful, visual publication about our nat-
ural treasures. The Atlas is now in its fourth
edition, with more than 40,000 copies al-
ready distributed; see [Chicago Wilderness
publ. |.
Mission-related benefits to individual mem-
bers. For the coalition to thrive, individual
members must become more successful i
their individual endeavors because of "e
association. with the coalition. Conversely,
each member must contribute in its own way
and with its unique strengths to the coalition
and to the pursuit of the collective mission.
For example, one of the goals of The Field
Museum is to direct its resources—scientific
innovative
expertise, worldwide collections,
education programs—to the immediate
Organizational chart for Chicago Wilderness.
needs of conservation at the re pona and i in-
ternational levels and to create a < be-
tween academic institutions and i groups
most closely tied to biological stewardship.
Chicago Wilderness offers the Museum a tre-
mendous advantage in tapping into a wide.
with Chicago Wil-
directed
regional effort. In turn.
the
more of its efforts in research, inventory, and
derness, Museum has much
tools for conservation to the Chicago region.
The potential for Chicago Wilderness to serve as
a model for urban conservation attracted early at-
tention of several federal agencies, including the
S.D.A. Forest Service, the U.S. Fish and Wildlife
Service, U.S.
Ageney, who have provided significant operating
and the Environmental Protection
grants since 1996, bringing new conservation rev-
enues to the region. Matching services from mem-
bers. and funds attracted by members for projects
catalyzed by Chicago Wilderness, exceed 70% of
the funds dedicated to this regional effort.
CHICAGO WILDERNESS: IMPACT ON SCIENTIFIC
RESEARCH
Research focusing on local biodiversity and land
management issues in the Chicago region has been
ongoing for over LOO years. But a recent resurgence
of interest in local issues by the region's scientific
community was coincidental with, and in large part
a consequence of, the founding of Chicago Wilder-
ness. This outcome can be attributed to several rea-
sons, including that the consortium has been suc-
cessful in bringing new resources to the area to
than the new
More
funds, however, is that the partnerships and oppor-
support research. important
156
Annals of the
Missouri Botanical Garden
tunities for collaboration fostered by Chicago Wil-
derness have created new research initiatives, iden-
tified important
interest in the resulting data, and created an en-
hanced atmosphere for the importance of biodiver-
research questions, expanded
sity and ecological research focused on the region.
Several Federal and Illinois State agencies have
provided grants to Chicago Wilderness to help en-
able the initiative to meet its goals. Chicago Wil-
derness makes requests for proposals each year to
distribute these funds to outstanding projects. A
proposal committee (see above) reviews submitted
proposals and ranks them based first on merit—
and then on the rel-
scientific, educational, policy
evance to the mission of Chicago Wilderness and
on the degree to which the results will make unique
contributions to the initiative; i.e., will serve as a
model for other projects or will provide data that
may serve as the underpinning for a series of other
studies. The topics of funded research proposals
fall into one of the following broad categories: Re-
search/Inventory/Monitoring; Restoration; and
Tools for Research—imaging projects, database
projects, or other conservation tools.
The following exemplars of science projects
funded through Chicago Wilderness provide an in-
dication of the breadth of the research undertaken
by Chicago Wilderness member institutions. All
projects funded through Chicago Wilderness are
summarized in www.chicagowilderness.org. Al-
though Chicago Wilderness funds only a small por-
tion of the flourishing biodiversity research activity
in the area, it has been greatly responsible for in-
vigorating the interests of scientists in conducting
research in the region.
RESEARCH/INVENTORY/MONITORING PROJECTS
Projects within this category focus on increasing
our understanding of the native biological diversity
in Chicago Wilderness. Research projects concen-
trate on factors influencing that diversity and on the
efficacy of current management practices to main-
tain and restore the region’s native diversity. Fund-
ed projects have ranged from studying the effects
of restoration efforts in oak woodland communities
on a broad group of taxa (fungi, beetles, millipedes,
amphibians, mosses, birds, and flowering plants).
to documenting the diversity and distribution of the
region's bats, to landscape-level inventories of wa-
tersheds, to studying the role of arbuscular mycor-
rhizal fungi in prairie restoration.
RESTORATION PROJECTS
The focus of projects in this category has been
(1) to develop best practices transferable to other
sites or (2) to restore specific sites of special sig-
nificance. Funded projects have ranged from de-
veloping and testing methods for restoring specific
threatened and endangered plants, to restoring a
degraded tallgrass prairie at Indiana Dunes Nation-
al Lakeshore.
TOOLS
The goals of these projects are to develop useful
tools that facilitate research and land management
activities in the region. These tools include imaging
projects to develop multispectral vegetation maps
using aerial flyovers, creating an online metadata
catalogue of existing spatial datasets owned by Chi-
cago Wilderness member institutions, and using
spectral data to track changes in land use and land
cover over the region. Funded database projects in-
clude compiling data from past breeding bird sur-
veys into a user-friendly database [Bird census] and
the development of a database of the current status
and location of the region's threatened and endan-
gered plant species [Chicago Wilderness grant]. An
example of another type of tool is the development
of a user-friendly computer model to analyze strat-
egies for deer removal in different land-manage-
ment programs [Ecology Software Server].
The continued growth and development of the
Chicago Wilderness coalition and the impact that
the coalition is having throughout the area, has en-
hanced communication among scientific research-
ers, land managers, educators, and policy makers
n the Such communication is critical for
effective development of research programs that en-
region.
hance the conservation of biodiversity. Through un-
derstanding the types of data most needed by the
users of research results (i.e., land managers, con-
servation biologists, policy makers, educators), and
how to present these data to the users, we have
identified new research projects while enhancing
the value of our research and expanding the audi-
ence for our results. In turn, the region has gained
better recognition of the value of research and has
increased opportunities for its continued funding.
The positive cycle continues to foster further re-
search activities in Chicago Wilderness.
OUR Roab Mar: THE REGIONAL BIODIVERSITY
RECOVERY PLAN
From the earliest days of the Chicago Wilderness
coalition, member organizations recognized that to
be effective, we needed a strong regional conser-
vation agenda with clear priorities for action, The
concept was to produce a biodiversity recovery
plan, in some ways analogous to a single species
Volume 89, Number 2
2002
Moskovits et al. 157
Chicago Wilderness
recovery plan, which would provide an assessment
of the current conditions along with a road map for
how to reach desired vision and goals. An effective
recovery plan would help shape decisions at both
the policy and the land-management levels.
We completed the plan in late 1999, after almost
three years of dedicated work by more than 200
members and nonmembers—including the science,
land management, education, and policy experts in
the region [Chicago Wilderness biodiversity |. The
Recovery Plan identifies the ecological communi-
ties of the greater Chicago region. assesses the
overall condition of these communities and the
threats facing each, lays out the vision and recovery
goals for each species assemblage. and recom-
mends actions to achieve our goals. The plan also
introduces a process of conservation and informa-
tion design that leads to coordinated research. in-
ventory, and monitoring agendas to support these
conservation goals and strategies and to anchor
them on the best available science. And the plan
calls for a massive communication effort to galva-
nize public support behind our conservation agen-
da.
ANATOMY OF THE RECOVERY PLAN
The plan provides both an overarching vision for
what Chicago Wilderness—the land and biodiver-
can become and a broad treatment of the de-
sily
tails of how to get there. In developing the plan. we
used a consensus-based approach, relying on the
region's experts and the available data. The plan
also provides the platform for organizing action at
a further level of detail, and it provides a strong
foundation for communication about the challenges
facing biodiversity conservation in the region.
The recovery plan document itself is intended for
reference. It is a comprehensive treatment of the
central issues, with 11 chapters and 11 appendices
Council. 1999
Chapters cover a summary of the biodiversity val-
(Chicago Region Biodiversity
ues in the region: a history of the changes in the
landscape and of conservation efforts to date: key
ecological processes that shape and sustain the re-
gions natural communities: the status. needs, and
goals of terrestrial and. aquatic communities: the
status of threatened species: research and monitor-
tools for conservation:
ing needs: goals for com-
munication and education: and the role for. key
players in executing the plan. To present the find-
ings of the plan to broader audiences, Chicago Wil-
derness developed а smaller, summary document
focusing on the important. issues and. recommen-
dations.
The recovery plan process brought together ex-
perts from many backgrounds and disciplines to
share their knowledge. The outcome is not only the
region’s state-of-the-art insights, but also a blending
of our science, policy, and communications exper-
tise. The process resulted in heightened interest
and commitment for collaboration in more detailed
plans for conservation action within a regional con-
ех!
The plan contains 140 separate rec ‘ommended
actions that fall into the following six categories:
(1
—
Land management. Most natural areas in the
region are not being managed in a manner
that will sustain plant and animal diversity
over time: more acres need to be under eco-
f
and the intensity (
-
logical management,
management must Increase.
Qe
М
2) Land protection. Additional land. will need
protection as the region’s human population
continues to grow. The primary concern in
setting aside more land for conservation is to
create larger, unfragmented protected areas,
either de novo or by enlarging existing ones.
Since much of the land in natural conditions
is either protected already or is in extremely
small parcels, the recovery plan. envisions
the need for restoring to ecological health
disturbed and converted areas, such as ag-
ricultural lands, as a strategy for creating
arger units.
мм
Stream protection, We must protect remaining
high-quality streams, where most of the re-
maining aquatic biodiversity resides, from
additional degradation.
Research needs
—
Research and monitoring.
relate primarily to understanding how best to
address the threats to biodiversity posed by
human development. For monitoring, the
plan puts forward a system of adaptive man-
agement based on conservation planning
(briefly discussed below).
(5) Education and communication. The plan rec-
ognizes that public understanding and sup-
port is a precondition to the success of
long-term, broad-based public undertaking
such as biodiversity conservation. The plan
lays out steps to achieve increased under-
standing of the importance of biodiversity
takes
short- and long-term scales (see further dis-
and what il to conserve it on both
cussion, below).
6) Local and. regional development policies. The
—
region is at a critical time for developing pol-
=
icies that will control anticipated growth iı
158 Annals of the
Missouri Botanical Garden
Table 2. Vegetation communities in Chicago Wilderness: a consensus list developed in 1999 by members of C hicago
Wilderness Science and Land Management Teams and by staff of the regional Forest Preserve Districts.
Community classes (8) Communities (24)
Subcategories (55)
Forested communities Upland forest
Floodplain forest
Flatwoods
Woodland
Savanna communities
Sand savanna
Shrubland communities
Sand shrub
Prairies Fine- айша. aoil prairie
Sand prairie
Gravel prairie
Dolomite prairie
Wetland communities Marsh
Bog
Fen
Sedge meadow
Panne
Seeps and springs
Cliff communities Cliff
—
Lakeshore communities Akeshore
Cultural (human-dominated) Croplan
vegetation communities Unassociated growth
Tree plantation
Developed land
Fine-textured-soil savanna
Fine-textured-soil shrubland
Dry mesic, Mesic, Wet mesic
Wet mesic, Wet
Northern, Sand
Dry m Mesic,
Dry mesic,
esic, Wet mesic
', Wet mesic
Dry, Dry —
Dry mesic, Wet mesic
Dry mesic, Wet mesic
Basin, Streamside
Graminoid, Low shrub, Forested
Calcareous floating mat, Graminoid, Forested
Neutral, Calcareous, Aci
Eroding bluff, Dolomite cliff
Beach, Foredune, High dune
Forb, Shrub, Tree
Grass,
population and in urban space in the coming
decades. The recovery plan stresses the im-
pact of the region’s management of growth on
biodiversity and suggests steps to mitigate
that impact.
THE PROCESS OF CREATING THE RECOVERY PLAN
The primary units of conservation in Chicago
Wilderness are plant communities. Initial natural
area inventories for the region, upon which most
conservation strategies to date have been based, fo-
high-quality communities
(White, 1978). The recovery plan also adopts plant
cused on remnant
communities as its primary conservation targets,
ut it considers animal communities or assemblag-
es, aquatic communities, and threatened species as
well. A critical first step in the plan became to
develop and adopt a plant community classification
system that all Chicago Wilderness members could
accept and upon which the region would base sub-
sequent conservation and land management deci-
sions (Table 2).
ost land management agencies in the region
traditionally have classified their holdings accord-
ing to the community classification system used by
the Illinois Natural Areas Inventory (INAI) in the
1970s (White, 1978). Rather than adopt a com-
pletely new system for the recovery plan, Chicago
Wilderness elected to develop a modified version
of the INAI system. The modifications reflect a
deeper current understanding of community gradi-
ents in fire-dependent forested lands in the region
(see Bowles & McBride, 1996). In addition. the new
classification places more focus on degraded lands,
recognizing those lands as the primary target of res-
toration efforts. While the Chicago Wilderness clas-
sification system does not follow the emerging na-
tional standard for community
(Grossman et al.. 1998), the recovery plan includes
a crosswalk to these national standards. The resul-
tant community classification system includes 8
major community groups, 24 community types, and
55 Ар distinguished primarily by soil mois-
ture (Table
While i * lays out actions covering all of
the biodiversity of the region, not only what
classification
is
unique or rare, it also incorporates a method to
identify what is most important from a global per-
spective and what conservation targets need the
most immediate attention. We undertook this eval-
uation primarily at the natural community level, but
also included information from key groups of ani-
mal assemblages and of threatened species.
Volume 89, Number 2
2002
Moskovits et al. 159
Chicago Wilderness
Regional experts first evaluated the status of an-
imal groups or subsets of groups: reptiles and am-
phibians. breeding birds. mammals, and some ter-
restrial invertebrates. In workshops, the scientists
grouped the species found in the region into broad
assemblages associated with habitat types: for ex-
ample. savanna birds or forest and woodland rep-
tiles and. amphibians. These assemblages are not
mutually exclusive in that some species will fall
into more than one grouping. The experts then as-
sessed the significance of the Chicago Wilderness
region to the global survival of each assemblage
and its constituent species. The three ranks we
used were: Globally Critical, meaning that the Chi-
cago Wilderness region is very important to the
overall survival and. well-being of the assemblage:
Globally Important, meaning that the region con-
tributes significantly to the overall well-being of the
assemblage: and Locally Important, meaning that
the assemblage is an important component of the
region's biodiversity, but its status in the region will
not have a substantial impact on the global well-
being of the assemblage. Experts also assessed the
condition of the assemblage in the region. ranking
3
them as “of concern/declining” or “stable.” Com-
bining the rank in importance on the one hand, and
the condition of the assemblage on the other.
helped to identify the Chicago Wilderness habitats
that need most immediate attention to stem de-
clines or to improve conditions. The exercise also
helped us to determine the habitats in the region
that are most significant in terms of global conser-
vation. The moist grassland bird assemblage came
out as both. Globally Critical and in poor condition.
Ten other assemblages in Chicago Wilderness are
Globally Important and are declining. АП of these
assemblages and conservation targets are listed in
the Biodiversity Recovery Plan.
Prior to evaluating communities, planners also
reviewed the threatened plant species in the region
and placed them into six categories of priorities,
from species that are globally rare to species that
may be adequately protected or stable. but are re-
stricted to communities that are rare within Chicago
Wilderness. Presence of threatened plant species
was an element in evaluating priorities for conser-
vation targets within biological communities.
To evaluate the relative need of different biolog-
ical communities for additional conservation action,
Chicago Wilderness members built on existing
methods for ranking endangered ecosystems (Noss
et al.. 1905: Noss & Peters. 1995). We modified
these systems slightly,
condition (quality), which is one of the most sig-
nificant elements in the sustainability of biodiver-
Table 3. Conservation targets for recovery based on
status, importance, and distribution.
to include the key factor of
(highest) tic
— (all i moisture classes)
Fine-textured-soil savanna (all moisture classes)
vanna
Sand prairie (all moisture gradients in dune and swale
lesic sand sa
topography)
Dolomite prairie (all)
Panne
Graminoid fen
Fine texture 'd-soil prairie (all moisture classes)
Second t
Dry sa ind savanna
ravel prairie (all)
Bania i marsh
Calcareous floating mal
Calc 'areous seep
Sand prairie xm than those in dune and swale
topogr raphy
Northern flatwood
Streamside marsh
sity in this urban region. As with the animal as-
semblages. expert groups convened for one-day
workshops over a period of several months to assess
available data and to rank natural community types
against defined criteria. Four rankings—quantity,
quality, biological importance, and contribution to
global conservation—combined to make the final
ranking. The ranking variables at focus were:
For quantity—(i) total amount (acres) remain-
ing of the community, (ii) amount lost. since
(111)
arge enough to sustain a relatively intact as-
pre-settlement, and number of patches
semblage of constituent species over time.
e For quality—(i) remaining percent in good
quality, (ii) degree of fragmentation and isola-
tion. and (iii) extent and effectiveness of cur-
rent management efforts.
For biological importance—(i) levels of species
richness, (ii) numbers of threatened species.
and (iii) key ecological processes.
For contribution to global conservation—rank-
ing of communities as (i) endemic to the re-
gion. (11) at the center of the range. or (iii) at
the edge of the range. This assessment also
considered the effectiveness of conservation ef-
forts for the community type in other regions.
We combined the individual rankings to develop
a tiered set of conservation targets, recognizing that
all community types will require some degree of
conservation effort to survive in the urban land-
Table 3 shows the first two tiers of commu-
nity targets. In all,
13 daylong workshops (in 1997—1999) for experts
scape.
Chicago Wilderness convened
160
Annals of the
Missouri Botanical Garden
to rank community types, and to discuss threats and
conservation needs. Through a similar process, we
evaluated the conservation needs of rivers, streams,
and natural lakes in the region.
The ranking system to produce the conservation
targets was not quantitative, but it represents a con-
sensus of the experts in the region. Key effort in
the planning process went toward developing and
adopting a process for further refining our conser-
vation goals, developing strategies to obtain those
goals, and setting up monitoring systems that en-
able information to feed back into refining manage-
ment strategies toward these goals.
The Chicago Wilderness region adopted a pro-
cess of conservation design, where planners (1) de-
velop a vision and goals for each conservation tar-
get, (2) identify key threats to achieving the vision/
goals, (3) set conservation strategies to abate the
threats, (4) establish indicators for progress toward
achieving the goals, and (5) develop a feedback
mechanism to adjust management strategies if mon-
itoring shows progress is insufficient. Chicago Wil-
derness members are using additional consensus-
based expert workshops to develop this iterative
process for conservation targets at individual sites
and across the entire region. Linking conservation
design to efforts that involve citizens in collection
of scientific data will help combine the scientific
aspects of the recovery plan with education, out-
reach, and public involvement programs.
Once assembled, the draft of the recovery. plan
went through a formal, legal hearing process under
the auspices of the Northeastern Illinois Planning
Commission and the Northwestern Indiana Region-
al Planning Commission. Both Commissions and
several municipal agencies adopted the plan within
months of its publication. The value of the formal
process was not only to gain feedback from a larger
public, but also to seize on the opportunity to raise
the issue of biodiversity in fora traditionally con-
cerned primarily or exclusively with development
and transportation issues. The final recovery plan
is now on a level with other major plans that will
guide the development of the region in the years to
come. As a result, the issue of biodiversity will be
discussed with business and civic leaders along
with economic competitiveness, quality of life, and
education.
OUR EDUCATION GOALS
Chicago Wilderness realizes that the future of
our native landscapes depends on the support and
involvement of our citizenry. On the encouraging
side, Gallup surveys (1999) and research by Belden
and Russonello (1996) reported that most Ameri-
cans consider environmental protection to be a pri-
ority. Concern about air and water pollution is
widespread; citizens worry about destruction of
tropical rain forests, toxic waste contamination, and
habitat loss. But while concern about these issues
is wide, it is not deep. Environmental issues often
fade when economic matters compete. And biodi-
versity recovery faces additional challenges: public
surveys report that only 20% of the population rec-
ognizes the term “biological diversity” (The Biodi-
versity Project, 1999)
Faced with these challenges, the Education and
Communication Team of Chicago Wilderness has
adopted (1) short-term communication strategies
coupled with active public involvement projects,
while (2) building and implementing longer-term
educational strategies. To guide our actions, the
more than 200 educators in Chicago Wilderness
have embraced the internationally held definition
of Environmental Education:
“Environmental Education is a learning process that in-
creases people's knowledge and awareness about the
environment and associated challenges, develops the
s and expertise to address these chal-
attitudes, motivations, i com-
mitments to make informed decisions and take respon-
sible action” (UNESCO 1978: 5: highlights added).
necessary
lenges, and fosters
The boldfaced words (our emphasis) spell out the
five components essential for making a program in
biodiversity education result in changed behavior
and in informed decision-making. Chicago Wilder-
ness Educators have used these five components—
awareness, knowledge, attitudes, skills, and partic-
ipation—to develop a matrix that assesses our ex-
isting programs, from birth to death, against our
targeted objectives.
Table 4 presents an abbreviated sample of this
education matrix. An “X” indicates presence of the
component in that program. For example, the BETA
project incorporates all five essential components.
Our compilation revealed that while our programs
are strong in awareness and knowledge, they are
weak or altogether lacking in attitude clarification,
skill development, and active participation, To ad-
dress these gaps and to reach the 80% who have
never heard of biological diversity, while also en-
suring that our education systems produce biolog-
ically literate graduates, we adopted program
guidelines that (1) extend the scope of our existing,
successful programs to include attitude, skills, and
action components; (2) develop new programs that
specifically address the lacking components; (3)
reach beyond our current circles by working with
community-based organizations; (4) expand the dis-
Volume 89, Number 2
2002
Moskovits et al. 161
Chicago Wilderness
Table 4
in that program.
Sample matrix of existing programs in conservation education. An “X” indicates presence of the component
Target audiences Existing resources
Essential components of biodiversity education
Awareness Knowledge Attitude Skills Participation
Children (Pre-School-8th Ап Atlas of Biodiversity X X
v Chicago Tribune Biodiversity Supple- X X X X
ment
Biodiversity Education Through Ac- X X X X X
tion (BETA) Project
Chicago Wilderness Communication X X
Tools
Mighty Acorns Expansion Program X X X X X
IDNR Biodiversity of Hlinois CD- X X X X
ROM Series
EcoWateh X X X X X
Adults An Atlas of Biodiversity X X
Chicago Wilderness Magazine X X
Chicago Wilderness Communication X X
Tools
EcoWatch X X X X X
Its Wild in Chicago N X X
tribution of our successful educational tools: and
(5) ensure that students in the Chicago Wilderness
region—from elementary through university—are
literate in. biodiversity.
Promoting programs jointly, creating complemen-
{агу (rather than competitive) projects, sharing
teacher-education expertise, and submitting collab-
oratively developed proposals to schools are some
of the new ways in which Chicago Wilderness
reaches adults. children, and educators for greater
impact. For example, Mighty Acorns is a steward-
ship program for eight- to ten-year-olds created by
The Nature Conservancy. This highly successful
program used to be limited to a few schools in one
county. Today, through the leadership of Chicago
Wilderness and two of its members (The Nature
Conservaney and The Field Museum). the program
has expanded to 19 Chicago Wilderness organiza-
tions in six counties in Illinois and one in Indiana,
and it involves 90 schools, 210 teachers, and 6500
students. The majority of these 6500 students come
from underserved communities. These elementary
students are engaged actively in learning about the
region's plants and animals, experiencing nature,
and practicing ecological stewardship—removing
buckthorn and garlic mustard, collecting and plant-
ing seeds—to improve the health of our natural
communities in the Chicago region.
Chicago Wilderness also has partnered with
large distribution networks, like the Chicago Tri-
bune’s Newspapers in the Classroom program, to
develop and promote a curriculum and activity
guide on the biodiversity of the region, using the
Chicago Wilderness Atlas as the central teaching
tool. Through its website and a printed brochure,
Chicago Wilderness provides to all in the region a
full list of the local educational resources and how
to obtain them [Chicago Wilderness publ.].
PUBLIC. INVOLVEMENT AND OUTREACH
To inform the public about pressing issues relat-
ed to the protection of our natural communities and
to engage a large constituency in the support and
work of restoration, the Biodiversity Recovery Plan
outlines the following priorities:
* Increase the public's understanding of the role
of ecological management
* Foster grassroots involvement in restoration
and conservation activities
* Expand opportunities for “citizen science
* Engage a wide public as volunteers
One opportunity that fosters wide public involve-
ment is an annual spring event, “Ws Wild in Chi-
cago." '
visitors to interact with Chicago Wilderness mem-
Ihe four-day festival attracts thousands of
bers at their displays through featured performanc-
es, hands-on activities, and targeted programs. The
festival provides a forum for the public to learn
about Chicago's wilderness and for Chicago Wil-
derness to reach new audiences with our messages.
162
Annals of the
Missouri Botanical Garden
Our National Public
effort fostered by Chicago Wilderness, enhances
Lands Day event,
family involvement as volunteers and the engage-
ment of a wide public in exploring the role of eco-
logical management in the survival of our native
communities. In 2000, more than 6000 people par-
ticipated in restoration activities at 21 Chicago Wil-
derness sites.
Through a program of the Illinois Department of
8 prog |
Natural Resources (IDNR; [EcoWatch
of citizens
). hundreds
adults and students—
high-school
gather data on the ecological condition of rivers,
forests, wetlands, and prairies in Chicago Wilder-
ness. Several more hundred “citizen scientists”
gather targeted data on birds and butterflies in the
instances, the volunteers. them-
region. In. many
selves enter the data on-line, to be analyzed by
regional scientists and land managers, and by the
state. Through our process of conservation design,
we are now tying the data collection by citizens to
the needs of land managers in monitoring and re-
vising their management efforts.
As a result of Chicago Wilderness, Field Muse-
um scientists and educators partnered with IDNR
to develop [UrbanWatch] with the following goals:
* Record the presence and distribution of groups
of organisms—plants, animals, fungi—in an
urban environment (we selected organisms that
can be used to assess the condition of natural
areas in urban centers)
* Examine environmental factors that influence
the biodiversity in urban ecosystems
Analyze and compile the information in a way
that is readily usable by urban planners and
park managers.
Focused on city parks, cemeteries, golf courses,
corporate Campuses, backyards, and
tree-lined streets, UrbanWatch collects data to ex-
schoolyards,
amine environmental factors that influence the na-
tive diversity in urban green space. Urban plan-
ners, park districts, site managers, and landowners
will have access to these data and will apply them
i
1 the management of sites to improve or create
urban spaces that are friendly to native communi-
ties and that address the needs of migrating birds
and butterflies in our region.
OUR COMMUNICATION OPPORTUNITIES
Significant current challenges in ecological res-
toration require immediate strategies to increase
public awareness and understanding. Chicago Wil-
derness has complemented its long-term approach-
es in public involvement, with actions aimed
another
achieving the following short-term goals for com-
municating about biodiversity to the general public:
* Recognize biodiversity in everyday experienc-
es
Understand human impacts on biodiversity
* Recognize the connection between healthy na-
ture and quality of life
The Chicago Wilderness Communications Team
has adopted the successful approach to crafting
messages developed by the Biodiversity Project
(1999) in Madison, Wisconsin.
vocales having a clear goal and target audience,
The approach ad-
identifying core values of the target audience that
the message embraces, and writing the message in
short paragraphs. These paragraphs address (i) why
the issue is important to the audience (leading with
values), (ii) what the threats are and who is re-
sponsible, and (iii) what people can do to address
the threats. Through this method, the message di-
rectly incorporates the values of the audience.
Equally important, each message ends with an ac-
tion that the target audience can easily undertake.
We developed a number of attractive and highly
effective communication tools using this method,
including a video, tabletop display, and slide show
[Chicago Wilderness]. These tools are available to
all Chicago Wilderness members, to deliver a con-
sistent, compelling message. Protecting Nature in
Your Community, a companion piece to the Biodi-
versity Recovery Plan, is one of our tools that fo-
cuses on reaching public officials, policy makers,
municipalities. zoning boards. and the over 300
governing bodies that review and guide the design
of hundreds of proposals for land development and
redevelopment every week. For the general public,
the Chicago Wilderness Magazine, currently with
over 9000 subscribers, is sold at local bookstores,
including Barnes € Noble and Borders, at gift
shops of our member institutions, and at other out-
els.
While the challenges to conservation education
and communication are many, the Chicago Wilder-
ness coalition enables us to work together in un-
derstanding our audiences, challenging our re-
sources, and creating novel approaches to engage
the widest public in maintaining and restoring our
local biodiversity.
Chicago Wilderness began with a larger-than-life
goal: nothing less than rescuing vibrant natural ar-
eas by transforming the environmental culture of
the people responsible for them. Through the part-
nerships formed and the collaborative work, Chi-
cago Wilderness is proving itself equal to the task.
From scientifically rigorous approaches in ecologi-
Volume 89, Number 2
2002
Moskovits et al. 163
Chicago Wilderness
cal inventory, restoration, and monitoring to crea-
tive advances in planning, policy, education, and
outreach, the accomplishments of Chicago Wilder-
ness are building a living legacy for the future.
Literature Cited
Be * ain and Russonello Research and Communic ations.
Human Values
Mute on Biological Diversity.
vello p search and Communications. ся D.C.
& J AcBride. 1996. Evaluation and
ina, Woodland. жк Barrens Nat-
The Morton Arboretum.
Bowles, шй
C iis ‘ation of Sani
Ju is in Northern Illinois.
asle, ili
i Biodiven TRIN —
Making the Connection:
tions J The Biodiversity
The Public.
Communica-
Madison.
1999. Life. Nature.
A Biodiversity
Project.
{ ‘onsi n.
C аз d Biodiversity Council. 1999, Biodiversity
Hec v Plan. Chicago Region. Biodiversity Counc il.
C hicago. Illinois.
Gallup News Service, the Gallup Organization. 1999. U.S.
public worries about toxic waste, air and Mer er pollution
as key environmental — CNN/USA Today/Gallup
Poll conducted March x
D. ж D. — tpe doen, А. S. Weakly. M
Bourgeron, R. Crawford. K. Goodin. S.
Төк K. D. Patters yr . Reid
1998. International Classification of
Ecological Communities: Terrestrial Vegetation of the
United States. Vol. I. The National Vegetation Classifi-
Grossman.
cation System: Development, Status, and Applications.
The Nature Conservancy. Arlington. Virginia.
Noss. et & R. L. Peters. 1995. —
Status Report on America’s Vanishing Habitat m
Defe nders of Wildlife, ш C.
LaRoe IH & J. M. Scot . Endangered
Ecosystems of the United States: | go — nary As-
sessment of Loss and Degri — Biological Report
28. National Biological Sen vice, U.S. Department of the
Interior, Washington
Sullivan, J. 1997. An Atlas of eas rsity. Chicago Re-
gion xr pis Council, Chicago, Illinois.
UNESCO. 1978. Final Report, ELE mtal Confer-
Environmental Education. ned y
UNESCO inc voperation t UNEP. Tbilisi. USSR.
26 October 1977. UNESCO ED/MD/19.
White, J. 1978. Illinois N vatural Areas Inventory Technical
Report. Illinois Natural Areas Inventory. Urbana. Hli-
nois
ecosystems
W шы
Internet Resour
[Bird census]: — //www. heldmuse 'um.org/birdeensus/)
[Chicago Wilderness|: (www.chic:
[Chicago W — ы ey (http//wwwe hic албаа.
r hin
[Chicago Wilderness grant]: (http://www.chicagowilderness.
— bom tml
[Chicago Wilderness publ. (http/www.chicagowilderness.
bi ations.htm
: (http: dl i state.il.us/or
: (http://nhsbig. im uiuc.edu/
асом ilde mess.org
ixl
p/inrin/ecc h)
software
ww/deer. — hin I).
[Urban We uch]: — fieldmuseum.org/urbanwatch)
№ ла ness Act]: (www.fs.fed.us/outernet/htnf/wildact.
htm
SAFEGUARDING SPECIES, Gary Paul Nabhan, Patrick Pynes,? and
LANGUAGES, AND кш
CULTURES IN THE TIME OF
DIVERSITY LOSS: FROM THE
COLORADO PLATEAU TO
GLOBAL HOTSPOTS!
ABSTRACT
Hotspots of — — have become priority areas for land conservation initiatives, oftentimes without recognition
that these areas are hotspots of cultural diversity as well. Using the Colorado Plateau ecoregion as a case study, this
inquiry (1) — the broad geographic patterns of biological ds and ethnolinguistic diversity within this ecore-
gion; (2) discusses why these two kinds of diversity are often influenced by the same geographic and historic factors;
and (3) suggests what can be done to integrate traditional ecological knowle dge of indigenous peoples into multicultural
conservation collaborations.
Key words: biodiversity, conservation planning, linguistic diversity, traditional ecological knowledge.
“Along come the sc ientists and make the words of our tablishment of Mesa Verde National Park in 1906
sto | ;
EURO пай and the Grand Canyon National Park in 191«
—Ag 1966 А
— (Burnham, 2000), more than 11.1 million acres of
"hs deir or de da — n — kanaat sa i ihe the Colorado Plateau's 130 million acres have been
diversity of human cultures. ... The intriguing ques- : . |
tion is this: apart from о md over resourc- federally protected for their natural and cultural re-
es, will the local communities bring back some of their sources. A diverse collection of national parks and
earlier cultural traditions of conservation of biological monuments, wildlife refuges, recreation areas, con-
diversity?
NIE _ servation areas, preserves, wilderness areas, and
—Gadgil (1987) : 1 . й
national historic parks апа sites, these protected
The Colorado Plateau of North America (Fig. 1)? lands are managed by the National Park Service,
has received international recognition for nearly a the U.S. Fish and Wildlife Service, the U.S. Forest
- x . А Servire : а "Or i € € 3009 Ta.
century because of the pioneering efforts there to Service, and the Bureau of Land Management (Ta
formally protect its spectacular natural and cultural bles 1 and 2). Conservation efforts on the Colorado
landscapes (Sellars, 1997). Despite that recogni- aleau were initiated long before our belated rec-
t $ 1- ` ars, . $ c : :
tion, the region's resource managers and conser ognition that the ecoregion harbors a remarkably
LI ^ р ` € D € D > Е : е ы
vationists have yet to work with much understand- high diversity of plants, butterflies, tiger beetles,
ing of how biological and cultural diversity have and mammals compared to 109 other ecoregions of
€ € € ` €
interacted within this four-state area. Since the es- similar size in North America (Ricketts et al.,
are grateful for the support of the Arizona Community Foundation, the Town Creek Foundation, the Ottens
Foundation, and Agnese Haury, all of whom enabled this synthe sis. Our manuscript was greatly improved as a result
of comments from Vic toria Hollowell, our Fl. agstaff colleagues Larry Stevens and Tom Sisk, David Armstrong, and two
anonymous reviewers. The numerous participants in the NAU/Te Se | кана sponsored retreat, “Bridging Ecological Res-
toration des Linguistic Revitalization on Indian Reservations on the ( ps rado Plateau," have greatly influenced our
thinking. In particular, we thank Vernon Masayesva, Ferrel Sekakaku, Luisa Maffi, David Harmon. and — Albert.
Lave * asayesva Jeanne and Donna House have also offered us guidance and insight into issues particular to the
Hopi and Navajo, respectively. Aimee Goodwin was оч м for GIS analyses in ARC-VIEW, and toge et with Dan
Boone and Ron Redsteer, kindly assisted us with ma
? Cente de or Sustainable Environments, Northern — University, Flagstaff, Arizona 86011,
" definition of the Colorado Plateau does not exist. ШЕ researchers
draw the — boundaries differently, but most agree that the Plateau's geographic heart is located in the Four
Corners region, where the states of Arizona, Utah, Colorado, and New Mexico intersect. The three — ine ін d in
this essay show two different boundaries of the Colorado Plateau, espec Чай, along the northeastern margin. For the
purposes of statistical analysis (of the total area of Indian lands versus federally protec ‘ted lands, ete.) in this dd we
used the ( Colorado Plateau. boundary outlined in Figure 3. The boundary shown in both Figures 1 and 2 is an “alter-
native? delineation.
..
s far as we are concerned. an objective
ANN. Missouni Bor. GARD. 89: 164-175. 2002.
Volume 89, Number 2 abhan et al. 165
2002 Safeguarding Species
dux AA - — · - =="
m.s Un М. UINTA MOUNTAINS I. Colorado
© | Salt Lake p —— m —— Yampa__ARiver
Ф ° / D
2 | { | $
L N | \ A 40 4
i N / $ \ т
à i
| | 483 | Na *
a Kilometers 134 ` 2
[. p C dM ; | © a
0 25 100 Y à | ue ] Y
; / $ WwW 2
A TS / y
| ё 2 /
А o | ak
2 y < С
| SCR І
K N
| | ^
Ё l | 7 38 +
| ABAJO MINS * Say,
| i
' V Lapara ^ MTS.
| ' v TS.
—X —
4 San Juan Rivet |
| | /
' / — * Y
T 20 E SAN JUAN BASN f
ig { we: Colorado | 8 l =
: ' | d
« Arizona | New Mexico $ І
b sf )
Бч с СА : Mt.
o, Flagstaff Ге) | dan I 6
< ' Ayr uerque
Pio, | l 1 |
120 110 100 =
| ip Y T Li
' 45 E
—
SONORAN | Ө
WHITE
DESERT
114 110 uns | 40 |
1 & =
as | Z I2
NC 7
Nilo 435
30 |
430
25
L L 1 DE
115 110 105 =н
Figure Boundary of the Colorado Plateau. For additional maps, articles, and photographs, see (http://
www.cpluhna.nau.edu/i ndex.htm).
1999a, 1999p). Although recent National Monu-
ment designations such as the Grand Staircase/Es-
calante have, in fact, taken into account the area's
biodiversity and rich cultural heritage, these two
factors have rarely been conceptually linked. More
typically, they have been offered as "twin" attrac-
tions for ecotourists intrigued by redrock land-
scapes. Indeed, the 8 and 10 million tourists who
annually visit national parks and monuments on the
Plateau may receive some unanticipated exposure
to this biodiversity and the ancient cultural influ-
ences upon it, but that is seldom what attracted
166
Annals of the
Missouri Botanical Garden
Table 1.
National parks of the Colorado Plateau. Sourc
e: National Park Service website (http://www.nps.gov)
Park State Acreage
Arches Utah 76,518
Black Canyon of the Gunnison Colorado 27,705
Bryce Utah 35,845
Canyonlands Utah 337,597
Capitol Reef Utah 241,904
Grand Canyon Arizona 1,217,403
Mesa Verde Colorado 52.121
Petrified Forest Arizona 3.53:
Zi ah 146,592
Zion
9 National Parks
2 in AZ; 2 in CO: 5 in UT
2,229.218 total acres
(
2.3 million acres)
them to the parks and monuments of the Painted
Desert or Canyonlands in the first place.
Because about 13.5 percent of the Colorado Pla-
teau’s landmass is already protected by federal
agencies, the ecoregion’s extant biodiversity has not
been considered as gravely imperiled as the bio-
diversity of other regions of North America. Nev-
ertheless, our recent (unpublished) survey of more
than 70 environmental professionals (including Na-
tive Americans) working on the Plateau indicates
that this ecoregion remains unusually vulnerable to
Table 2. National monuments of the Colorado Plateau.
agement websites (http://www.nps.gov) and (http://www.blm.
threats such as the damming of rivers; oil, gas, coal,
uranium, and aquifer mining; competition from in-
and the fragmentation of wildland habitats by rap-
idly increasing urbanization. Such threats continue
vasive species; mismanagement of wildfire regimes
to diminish native biodiversity, both within and be-
yond national parks. The Nature Conservancy
(TNC) has therefore ranked the Colorado Plateau
within the third tier of hotspots of imperiled bio-
diversity. At the same time, TNC recognized that
۰
this ecoregions “rarity-weighted species richness’
Source: Nalional Park Service and Bureau of Land Man-
gov.)
Le)
Name
State Managing agency Acreage
Aztec Ruins New Mexico NPS 31
Canyon de Chelly Arizona NPS (leased from Navajo Nation) 83,840
Canyons of the Ancients Colorado BLM 164,000
Cedar Breaks Itah NPS 6,154
Colorado Colorado NPS 20,453
Dinosaur J NPS 210,277
El Malpais New Mexico NPS 114,277
El Morro New Mexico NPS 1,278
Grand Canyon-Parashant Arizona BLM 1,014,000
Grand Staircase-Escalante Utah BLM 1,900,000
Hovenweep Utah NPS 784
Montezuma Castle Arizona NPS 857
Natural Bridges Utah NPS 7.036
Navajo Arizona NPS (leased from Navajo Nation) 360
Pipe Spring Arizona NPS 40
Rainbow Bridge Utah NPS 160
Sunset Crater Volcano Arizona NPS 3,040
Tuzigool Arizona NPS 10
Vermilion Cliffs Arizona BLM 293,000
Walnut Canyon Arizona NPS :
Wupatki Arizona NPS 35,422
Yucca House Colorado 33
22 National Monuments 10 in AZ: З in CO;
3in NM; 6 in UT
NPS
BLM and NPS lands
3,859,549 total acres
Volume 89, Number 2
2002
Nabhan et al.
Safeguarding Species
was considered more significant and less well-
known than it deserved to be (Stein et al.. 2000):
in other words. the Colorado Plateau harbors mans
biological rarities whose vulnerability to threats has
not yet been adequately assessed. Consider the fact
that the relatively well endowed Grand Canyon Na-
tional Park has continued to lose an average of one
species per year during the last two decades: the
park’s current research budget and resource man-
agement strategies have somehow not been. suffi-
cient to prevent the local extirpation of rare spe-
cies.
Putting aside for the moment the degree to which
current threats imperil the ecoregion's biota. it is
clear that the Colorado Plateau is indeed rich i
such rarities. including endemic species (that is.
species with narrow distributions that occur in this
ecoregion and nowhere else). Continent-wide floris-
Kartesz and Farstad (1999)
affirmed that the Plateau is the ecoregion of con-
пе analyses by have
tinental North America with the highest rate of vas-
cular plant endemism, reporting 290 species re-
For the
David Armstrong (in
stricted to this ecoregion. the fauna of
Colorado Plateau ecoregion,
prep.) has recently determined that 23.6 percent of
the mammals and 36 percent of the rodents exhibit
endemism at the levels of species or subspecies.
Vhile appreciation of the Plateaus biological
uniqueness has grown, recognition of its cultural
and linguistic uniqueness still lags far behind. The
Colorado Plateau is home to more speakers of Na-
tive American languages than all other regions in
the United States combined. The ecoregion’s indig-
enous peoples belong to 24 different tribes. bands.
or dialect communities and represent six different
language families (Table 3).* (English. of course, as
well as Spanish and Basque are also spoken here.)
Among the Plateau’s indigenous languages. Zuni is
a language isolate, or what biologists might call an
Ас-
(a:shiwi) has no
language of the Colorado Plateau.
Zuni
close relative in any other language family (Camp-
bell, 1997). In addition to Zuni, the other language
families indigenous to the Colorado Plateau include
“endemic”
cording to many linguists.
Keres. a family represented by Acoma Pueblo, La-
the
language family represented at Jemez Pueblo: Uto-
guna Pueblo, and Zia Pueblo; Kiowa-Tanoan.
Aztecan. the language family to which the Hopi.
Ute, Athabaskan.
represented by the Apache and Navajo languages:
and Paiute languages belong:
and Yuman. represented by the Yavapai, Havasu-
'The numerous sources for Table 3 are found in the
Literature Cited. and each source is preceded by an as-
terisk (*).
pai. and Hualapai tribes on the Colorado Plateau.
Figure 2 shows the approximate geographic bound-
aries of these six indigenous language families in
1850. when the Colorado Plateau officially became
part of U.S. territory. Tribes and communities who
spoke (speak) a language that belongs to one of the
six families were living in these areas at that point
in time. For example, Navajo and Apache speakers
occ upie d the lands shown as "Southern- Mhabas-
can" on this historical map.
The Colorado Plateau undoubtedly ranks among
the top five American regions north of the Tropic
of Cancer in terms of linguistic. cultural. and bio-
logical diversity. as well as in biological and lin-
guistic/cultural endemism. Nevertheless. there is
not a single conservation plan that takes into ac-
count both the cultural diversity and the biological
diversity of the region. It is as if the historic and
geographic relationships between “nature” and
“culture” on the Plateau are somehow irrelevant. o
too hotly debated to be of value in conservation
planning. While it may be reasonable for conser-
vation planners to be skeptical of painting all Na-
tive American land and water management practic-
es as “ecologically noble,” it is also tragic that so
few Native American communities have been in-
volved in planning national parks and monuments
adjacent to their current reservations. These federal
ands were clearly parts of their historic homelands
2000).
resources and the management of natural
(Burnham, As a result. the management of
cultural
resources have typically been done by different sets
of specialists, sometimes involving Native Ameri-
cans in the former but nearly always ignoring their
traditional ecological knowledge of the latter.
This historic failure of the vast majority of con-
servation biologists and environmentalists to sub-
stantively engage Native American communities in
collaborative work based on shared goals is both
the
residents of the Colorado Plateau have
disappointing and ironic. It is ironic because
long-term 1
substantial knowledge about the history of the local
flora and fauna that is not available from other
sources. Even if all their current hunting. foraging.
or farming practices are not considered to be eco-
logically benign by conservation. biologists (Dia-
mond, 1993). this does not negate the value of their
traditional ecological knowledge (as defined bs
Berkes.
animal distributions, densities, and vulnerabilities.
2000) about factors influencing plant and
[Incidentally. Diamonds (1993) widely cited con-
demnation of prehistoric peoples of the Colorado
P
obtain timber
ateau for deforesting the Chaco Canyon area to
build multi-storied pueblos has
been refuted by recent strontium isotope evidence
Annals of the
168
| Garden
issouri Botanica
M
591615 ]u313]]rp 1n0j ut
c9reee sor[rurej ozensue[ O sosensue[ QI ZESTZZO OZ вәшипшшо 10 *spueq “SIGLO JOLOSIP pz
209 2,99 0696 (s93Ane]9.1 UMOUY ou) типу iun Олс EOF o[q9ng tunz
E> AOL LL UESSI9N 8949M use 000 ZI I o[qonq BIZ
AOC 25€ ! €90'€ UPOUP? [-емогу PMOL 619°68 ojqənd zəwəf
Me ACO 009 UBOUB | -PAOIN ema] (ema, ido) ouey
28> 081 обот BEY O] Чоң 90€ СРС o[qong idoy
We WOE 969"! ивѕәләҹ̧ вәләў UASA LE L6P o[q9ng випЗет
HT 250€ ТРО ивзәләӘ 5919 y ULISIA GOC BLE o[qang BWOOY
LE 0с 000€ Неее] эп 812909 онно зд
YEC> LEE 00€€ UE o] 91 000'00€ t 219 Хет) цеш
%G %SI 1 usen yon An B8C ELE ап Эрн)
2501 жс} 00€ U£29]ZV-017] әр шәціпос$ 000'€ 3Inieq urouinog uenf ueg
%0 % с> ESL mor AU шәшпос 000 0t Чел jo эди], этед
20 E> OVE AA Bie amed ureqinog LEPOTI вивтрц әйе Jo pueg чесе
%с> E> 8S I Heina Tampon) ed coe Hoosa14-redeaeq
Mg 2562 C/9'[ uexseqeuiy es x чешпд -ruyo эцоваұ Y ed 9€9 ayoed y -redeaex opio. dure?)
ec 750€ DOPZ AA TIOS ed 000'000' I redepeng
7506 7586 6t9 Еши IO ed 220'881 rednseae yy
25€ CL coe'c uexseqeui y -yey OÍBABN 000'08 (sofeAeN Jo pueg oj20ue7)) gyoopufeyoy,
20€ 7509 єс UP ASE qeu y -Jei ofPary ESOP I ofeAeN, yewey
200€ 251€ 9cc'occ ueyseqeury YEA ofeary 000'000' 2 1 uoney ofeaev
2506 25€6 000'c uByseqeyly-xery aqoedy 000'£9 ofesey oure[y
LEI 250€ 000'CT uexseqeqy -e43 ayoed y 000'009" 1 əyəedy urejunopy AYM
%0 250€ OL ueseqeuy YEA aqoedy cg auoedy ошо],
256 HET 00€'€ ueyseqeu y -xe4 ayoedy [09°68 auoedy ejpueorf
81—@ so3e siaxeads зләчшәш uonerpgje A ruej dnou3qns a3en3ue] (sə1ə2) Хиипшшоә 10 *pueq *aqup[
‘sraypads əaneu uan [equa adenzue] Jo 93en3ue"] 3seqpue[ [equ едо,
әлцви Juang jo a3ejuadiad po[[o1ua [eio]
jo a3ejuadiad ojeuixo1dd y
3jeuirxoadd y
(м) Х51ЛӘ15Р ue Aq popaoaud S1 991nos qoe] “UOLIS вәәпәлә}ә 34] ut UMOUS әле eyo sių} 10} S3JINOG :310N ‘sağen due] snouaZiput S,ne3]e[d оре1о[о? эч jo snes СС 3IqeL
Volume 89, Number 2
2002
Nabhan et al.
Safeguarding Species
169
е
€ 5
Via ou ENS
e foes TEGAN
SOUTHERN-ATHABASCAN
Locations of Indigenous Language Families
on the Colorado Plateau in 1850
GIS Support From
Figure 2.
y
190 Kilometers E
Locations of indigenous language families on
Center for Sustainable Environments
that the Chaco Anasazi obtained their ponderosa
and spruce/fir beams by cutting a single age class
of trees selected from two large mountain ranges
50-75 miles away (English et al
. 2001).| Indeed,
CANYON CONSERVATON TECHNOLOGIES
the Colorado Plateau in 1850. (From data developed by the
the more pertinent question is whether the tradi-
tional ecological knowledge of Plateau tribes is be-
ing lost at such a rapid rate that it will no longer
be of service in biological conservation (Nabhan
172
Annals of the
Missouri Botanical Garden
ligations (L. Masayesva Jeanne, in Nabhan $
Reichhardt, 1983).
Traditional ecological knowledge can also be
useful in locating and staging rare species reintro-
ductions and habitat restoration. At a time when
the (
were on the verge of failure, Rea (1981) proposed
that the recovery team broaden its view and con-
sider reintroductions in the historic refugium near
the Colorado River's Grand Canyon, where oral his-
tories from the Hopi recall sightings of the giant
call Awaatoko (Whiting unpublished, in
1993). This has, in fact, turned out to be
one of the more successful areas for Condor rein-
;alifornia Condor recovery efforts in California
bird they
Seaman,
troduction, even though it is outside of California
where many biologists erroneously presumed that
the California Condor belonged (Rea, 1981).
INVESTMENT IN CONSERVATION
CAPACITY-BUILDING FOR NATIVE
REMAINS CRITICAL
AMERICANS
For decades, the Bureau of Indian Affairs hardly
involved Native Americans in so-called technical
decisions regarding biological resource and land
management options on tribal lands. While the
number of Native Americans trained in biology and
nature resource management increased fivefold be-
tween 1970 and 1999, there remains a chronic un-
derinvestment in training Native American profes-
sionals in biodiversity conservation on tribal lands
relative to the training available for professionals
managing federal lands. The Navajo Natural Heri-
tage Program, first sponsored by The Nature Con-
servancy, has become an outstanding example of
the "conservation -off" of such investment
(Stein et al., 2000). The Hopi, Zuni, and White
Mountain Apache tribes have also developed pro-
grams that include wildlife management, endan-
gered species recovery, and ecological restoration.
In the case of the Zuni, their program to reintro-
duce beavers as part of the tribe's riparian resto-
ration efforts has involved elders teaching youth
about the cultural significance of beavers and other
wetland dwellers (Albert & Trimble, 2000). When
both Western academic and traditional ecological
knowledge are valued by such programs, it ensures
that Indian youth interested in natural history are
not forced to choose between their own cultural way
of looking at the natural world, and the way of mod-
ern science. One superlative example of such ca-
pacity-building efforts among Native Americans is
the EPA-funded Institute for Tribal Environmental
Professionals based at Northern Arizona University,
which has already trained more than 500 tribal
members from over 200 communities in inventory,
monitoring, and restoration techniques for environ-
mental protection.
MORE LINKAGES ARE NEEDED BETWEEN
ECOLOGICAL RESTORATION AND LINGUISTIC
REVITALIZATION
In addition to the many ecological restoration
projects recently initiated on tribal lands, most Na-
tive American communities on the Colorado Pla-
teau are actively engaged in language stabilization
1996). Т
include
hese
language the
elaboration of the entire extant lexicon of a tribe,
and revitalization efforts (Cantoni,
restoration projects often
including its names for plants, animals, and their
habitats. In eliciting these names, linguists often do
not know the particular scientific taxa the words are
describing, but nevertheless record. ecological in-
formation that may be of use to ecological restora-
tion. and recovery projects (Nabhan,
2000a). For instance, the Diné name for the Indian
paintbrush, Castilleja lanata A. Gray, is dah yütih-
"big hummingbird's food," to
spec ies
idaa tsoh, meaning
distinguish it from the beardtongue, Penstemon bar-
batus (Cav.) Roth, called dah yiitihidaa'ts‘ooz, or
“food of the slender hummingbird.” It may be that
Diné elders recognize that these two flowers are
frequented by different sizes and species of hum-
mingbirds, and that such information can be inte-
grated into ecological restoration efforts to increase
=
forest understory biodiversity.
CONCLUSIONS: IMPLICATIONS FOR GLOBAL
CONSERVATION STRATEGIES
Over the last decade, tremendous effort has gone
into defining, locating, and rapidly assessing the
hotspots of biodiversity” (Mit-
1998;
these
species richness of *
termeier et al., 1998; Olson & Dinerstein,
Ricketts et al., 1999a).
hotspots not only capture a large proportion of the
planet's biodiversity, but they also define an agenda
By some accounts,
for conservation investment, which is largely ex-
pended on land purchases and infrastructure de-
ve opm nt for protected areas (Mittermeier et al.,
38). By investing an average of $40 million/year
on land purchases in hotspots, rather than on the
—
current “scattergun” approach of current conser-
vation expenditures, these conservation strategists
have proposed a new “silver bullet” to biodiversity
loss that could protect areas containing a high pro-
portion of the world’s species, while business out-
side the hotspots could be allowed to go on as usu-
^
mE
—
—
Ehrenfeld (in press) has pointed out the serious
170 Annals of the
Missouri Botanical Garden
2000a). To date, this rapidly disappearing orally Americans on the Colorado Plateau have vet to re-
transmitted knowledge has rarely been systemati-
cally taken into account and valued by land man-
agers, who often consider such knowledge to be the
quaint but anecdotal or superstitious recollections
of scientifically untrained old-timers.
Even contemporary Native Americans’ roles in
managing the lands of the Colorado Plateau have
been chronically underestimated. While the Grand
Canyon Trust considers “Indian country” to com-
prise just a quarter of the Colorado Plateau’s 130
1996: Wilkinson,
1999), our GIS-based analyses more accurately es-
tablish that 29-32% of the Colorado Plateau is be-
ing managed by Native American communities (de-
million acres (Hecox € Ack,
ending upon which definition of the Plateaus
| 8 Ul
t
would be presumptuous if not impossible to develop
boundaries is used). By virtue of this fact alone,
a systematic conservation plan for this ecoregion
that did not take into account Native
land stewardship and traditional ecological knowl-
American
edge of endangered species on the Colorado Pla-
teau (Nabhan, 2000a, b).
In a modest effort to begin to bridge the historic
gap between studies of biological and cultural di-
versity, Northern
Sustainable Environments (CSE) has initiated a pi-
Arizona University’s Center for
lot study with the Grand Canyons Wildlands Coun-
cil to assess ways to better safeguard and restore
both the biological and the cultural uniqueness of
the Colorado Plateau. In designing this pilot study,
we have investigated the potential linkages between
biological and linguistic diversity elucidated by the
scholars involved in “Terralingua: Partnerships for
Linguistic and Biological Diversity,” an internation-
al non-profit organization hosted by CSE, which has
recently published a global analysis of biocultural
(Maffi, 2001).
emphasizes why efforts to conserve biological and
diversity The following discussion
linguistic diversity should be linked, whenever pos-
sible, using examples from the Colorado Plateau.
BIODIVERSITY ON NATIVE AMERICAN LANDS
CONSERVATION INVESTMENT
THE
DESERVES MORE
Indian reservations in the United States cover
tens of millions of acres of the North American con-
tinent, comparable in extent to the acreage that the
National Park Service manages for North America’s
biodiversity. In particular, reservation lands on the
Colorado Plateau (Fig. 3) cover nearly twice the
area that national parks, monuments, preserves.
conservation areas, wilderness areas, and wildlife
refuges cover in the same ecoregion. Nevertheless,
the 29.6 million acres of lands managed by Native
ceive much investment from federal or private
sources for the inventory of their biodiversity, mon-
itoring, and recovery of their rare species, or local
capacity-building in environmental protection—
relative to the considerable support given to those
working on adjacent federally protected lands.
If all the species found on Indian lands were also
found on lands rigorously protected by the National
'ark Service or The Nature Conservancy, perhaps
this issue would be easier for conservation biolo-
gists to ignore. However, the narrowly distributed
endemics of the region are often restricted to hab-
itats found only on reservation lands, and not on
park lands. The Navajo sedge, Carex specuicola J.
T. Howell, is an endangered species found around
only three springs and seeps used by Diné (Navajo)
livestock herders, and one spring used by Hopi
farmers. In other words, its range is restricted to
the Navajo and Hopi Reservations (D. House, pers.
comm.; Nabhan et al., 1991). Similar situations oc-
cur for the endemic Hopi chipmunk (Tamias rufus).
a subspecies of the Spotted Ground Squirrel (Sper-
mophilus spilosoma cryptospilotus). the Chuska and
unitcha Mountain subspecies of Abert's squirrel
(Sciurus aberti chuskensis), and a subspecies of Ste-
phen’s woodrat (Neotoma stephensi relicta) found
only on Navajo lands. Even if conservation biolo-
gists continue to feel ill-equipped to deal with the
cultural and legal (sovereignty) issues regarding
plants and animals restricted to tribal lands, they
can no longer ignore the fact that the only means
to sustain these species is by providing Native
American land managers with the resources needed
lo protect or recover these rare populations and
their habitats.
CONSERVATION COMMUNITY NEEDS TO
TRADITIONAL ECOLOGICAL KNOWLEDGE
THE
EMBRACE
Until recently, endangered species recovery
teams and ecological restorationists disregarded the
Native
Perhaps the first formal
traditional ecological knowledge found
American communities.
breakthrough occurred when Diné biologist Donna
House incorporated traditional ecological and eth-
nobiological knowledge about the Navajo sedge into
its federal species recovery plan (House in Nabhan
et al., 1991), acknowledging that Diné herders had
been stewards of this plants habitat for decades
further efforts to
and deserved to participate ii
safeguard the sedge.
In addition to their role as habitat managers,
some cultural communities on the Colorado Plateau
intentionally protect rare species, as is the case
Volume 89, Number 2
2002
Nabhan et al.
Safeguarding Species
t
%
=
— z
2 XK ©
Sr A =
/ \ % e
Northern Ute / s N * è
oA р p, 1 \ $ A d ~
k OR сусту"
\ T.
Oph ute Р \
' ۱
| *
| [фр
) ee
/ /
/ |
4 е Grand Junctíon
Paiute / е
FI X |
L p A. E
24 N
)
f А
/ E
/ Ё
* Ж E
Southern’
A
Paiute, Ute Mountain |
Southern д ; \
( Southern Ute
Paiute outhern Durango g^
| — m ha 3 Jicarilla Apache
5: Cur | —
y SER
\ А
—
Г
Havasupai
/
°
\ Flagstaff
Prescott Camp Verde
$ * \
|
\
Native American Indian Reservations ~~
on or near the Colorado Plateau
Native American
Reservation
p Colorado Plateau
5 Bound
Figure 3.
Native American Indian Reservations on or near the Colorado Plateau.
with the anomalous sunflower (Helianthus anomalus tices (Nabhan & Reichhardt, 1983). Its flowers may
S. F. Blake). Of the twenty-five known localities for be the sole source of a ceremonial facepaint pre-
this species, at least seven are in or adjacent to pared for the Lakon mana (maiden) ceremonies of
Hopi fields and kivas, where farmers and priests early fall, so that these sunflower populations are
spare it from their otherwise rigorous weeding prac- protected as a necessity for fulfilling spiritual ob-
Volume 89, Number 2 Nabhan et al. 173
2002 Safeguarding Species
flaws associated with such an agenda, the most se- lands held in trust for future generations of in-
rious being that most hotspots are not only inhab- digenous peoples (Burnham, 2000).
ited by diverse cultures, but they are also not for œ It is — that for such high levels of biodi-
sale. By the mid 1990s prominent ethnobiologists, versity to persist in any hotspot until this day,
anthropologists, linguists, and biogeographers had —— residents consciously or unconscious-
brought attention to the fact that the 10-20 richest ly developed active and passive means of man-
biodiversity hotspots were also extremely rich in aging particular landscapes, microhabitats, or
cultural diversity, which was typically defined in species guilds to maintain them. The potential
terms of the richness of extant indigenous languag- utility of such knowledge, skills. and practices is
es (Harmon. 1995). For instance, Toledo (1994) ob- great (Nabhan, 2000b), so much so that species
served that of the nine countries considered to have recovery teams and formal land managers should
the highest species richness of mammals, six of be encouraged to integrate such knowledge into
these were also among the richest in indigenous their conservation management plans (Tuxill &
language groups: Indonesia, Brazil, Mexico, Zaire, Nabhan, 1998).
India, and Australia. e It is likely that proponents of biodiversity con-
Three commentaries are typically offered to ex- servation will forge stronger, more effective col-
plain this correlation and to hypothesize its causes laborations with indigenous residents in biodi-
(as cited by various authors in Maffi, 2001): versity hotspots if they listen to and respect
indigenous concerns about sovereignty, cultural
* Geographic determinism: Both species and lan- property rights, and secrets associated with eso-
guages diversify in heterogeneous. landscapes LUPIS aul — —
c У. » f H sts
with formidable geographic barriers.
: . ‘ a af A EOM "EET should attempt to collaborate on community-
* Ecological determinism: Linguistic diversification DUM ueste Qai Ner ihe асана bd
. . . H . “гг c 5 >C S c 5 > > M ? t
"n e “AR : o rere qe эү.
occurs in response to high biodive хаку, as differ transmission of traditional ecological knowledge,
е — фо — — T DM sss M indicis languages, eds
niches for themselves and encode their knowl-
| Frich bi НТ, tional subsistence or ceremonial practices, and
edge of rich biotas in different ways. ; :
— "ir ea P SS the generation of health benefits or income for a
* Historic determinism: Areas of high linguistic di- eh АВ i Жолини ашин”
, ; с e Oi ross section of co / members.
versity today are "residual," persisting because We i m i
"ede. e ee — uc uw While such community-based conservation efforts
of their geographical isolation from extensive ag- BE e lta asia fs | |
жент! devel а, 1 are already given lip service from governments
ricultural development (implying that now-impov- ~ І i І .
ished edi and NGOs alike. there remains a disproportionate
erished areas were once more diverse). ;
investment in “top-down” conservation strategies
Clearly, our Colorado Plateau pilot study reiterates and only scattered. investment in community-
a pattern seen on other continents as well: where based “bottom-up” strategies for maintaining bio-
hotspots of biodiversity or endemism are inhabited diversity.
by a diversity of cultures, the ways in which these * Given the fact that traditional ecological knowl-
cultures encode traditional ecological knowledge edge about landscape-level biodiversity itself i
about species in their native languages has tremen- being diminished (Berkes, 2000), the fragile re-
dous potential for helping to conserve this biodi- lationship between the two needs to be more tan-
versity (Nabhan, 2000b). From this frequently ob- gibly explored. While many biogeographers and
served correlation, several management linguists are already involved in efforts to under-
implications must be considered: stand global patterns of biodiversity and cultural
diversity, it is hoped that our pilot project to safe-
e It may not be feasible or even advisable for gov- guard the uniqueness of the Colorado Plateau will
ernment or non-government organizations (NGOs) encourage other community-based. practitioners
to purchase the lands within these hotspots for to experiment with more practical MAN ION
biodiversity protection, because much of the bi- tegrating indigenous knowledge into collaborative
ologically diverse area may be “common lands” efforts to conserve the world’s hotspots of diver-
held in trust by these cultural communities, such sity and endemism.
that they cannot be purchased, traded, or con-
demned. New efforts to conserve. biodiversity
elsewhere in the world need not make the same Literature Cited
Latera + Naton ork arias Ha do TEN Е
mistake the National Park а historically Agnon, S. Y. 1966. Editorial commentary. The Observer
made on the Colorado Plateau, by stealing, con- 5(1): 6.
demning. or finagling the rights to species-rich Albert, S. & T. Trimble. 2000. Beavers are partners in
174
Annals of the
Missouri Botanical Garden
riparian restoration. on ed — Indian reservation.
Ecol. Restoration 18(2): €
Armstrong, . in prep. е. in mammals of the
Colorado Plateau. Draft analysis in preparation by the
author, Department of Environmental, Population and
Organism Biology, University of Colorado, Boulder, Col-
orado.
*Benson, — dad o of Utah.
view. 12 Decembe
— 2000. el Ecology
Telephone inter-
Yale Univ. Press, New
— P. 2000. Indian Country, God's Country: Native
Americans and the National Parks. Island Press, Wash-
ingtor
*Campbe ll, т, 1997. American Indian Languages: The
Historical ee of Native America. Oxford Univ.
Press, New
Cantoni, G. (editor). 1996. Stabilizing Indigenous Lan-
guages. Northern Arizona University Center for Excel-
lence in Education, Monograph Series, Flagstaff, Ari-
zona.
*Cuch, Mariah. Ute Tribe Public Relations. P.O. Box 400,
Ft. Duchesne, Utah 84026, Electronic mail correspon-
dence, 27, 29 December 2000.
оо J. 1993. The Third Chimpanzee. W. W. Norton,
or
Ehrenfeld. In press. Hot Spots and the Globalization
of coran Oxford Univ. Press, te ford
tu . Betancourt, J. 5. Dean & J. Quade.
001. Ае isotopes reveal d sources of ar-
chitectural timbe sr in Chaco Canyon, New
а! . Acad. Sei. 98: 11891-11896
. Wilfred. — red Zuni Public
. Personal intery November 2000. Zuni
; Mexico.
*Estes, James. October 1999. “How Many Indigenous
Languages Are Spoken in the United States? By How
Many Speakers?" National Clearinghouse for Bilingual
Education. (htt tp://www.ncbe.gwu.edu/asknebe/faqs/
20natlang. epe
Gadgil, M. 1987. шо,
Trends EU Evol. 2(12): :
*Gallegos, Estrella. Sia nd 2 anguage De —
Ute Mountain Ute Tribe. Telephone interview. 1 Noven
ber 2000
=
=
Mexico. Proc.
ч and biological.
13.
*Glassco, Greg. Cultural Resources Office, Hualapai
2
Tribe. Peach Springs, Arizona. Telephone interview.
x
anguages of the World.
dated February 1999 at
TT
ل
org/e e le
йш, D. 1995. The status of the world’s languages n
ntes in ioe Southwest J. Linguistics 14:
2"
— W. К. Ack. 1996, Charting the Colorado
Plateau: An — ne De ‘mographic Exploration.
Grand —— Trust, Flagstaff, Arizona.
*Hussey, Jenni en es Direc tor, Paiute Tribe of Utah
— чанган, Cedar City, Utah. Telephone in-
terview. 20 De М.
— J
^cember 2
A. Farstad. 1999, Multi-scale analysis of
in-
— all web sites shown) that were used to construct
Table
—
*All references denoted with an asterisk are sources
1
endemism of vascular plant ege ies. Pp. 51-55 in T. H.
Ricketts, E. Dinerstein, D. M. Olson & C. Loucks, Ter-
restrial Ecoregions of North America: A Conservation
Assessment. Island Press, Washington, D.C.
*Krauss, Michael. 1996. Status of Native American Lan-
suage Endangerment. Pp. 16-21 in G. Cantoni —
Stabilizing Indigenous Languages. Center for el-
lene e in Education, Northern Arizona pop Flag.
aff.
Maffi, L. 2001. On Biocultural Diversity. Smithsonian Uni-
ity Press, Fie ass ee py a
Mittermeier, R. . Myers, . Thomsen, G. A. b.
Fonseca & S. ‘Olivieri 1998. Biodive 'rsity hotspots and
major tropical wilderness areas: Approac 3 ч
conservation priorities. Conservation Biol. 1 520.
Nabhan, G. Р, 2000a. Interspecific M. jene
endangered species recognized by sire and Com-
caac о ures. Ecol. Appl. 10: 1288—1295
2( The — Plateau as a center of «
versity. idend Newsletter, The ( Consultative Group
on bu dn Diversity, San Fran ncisco.
. L. Reichhardt. 1983. Hopi protection of He-
— avt x a rare sunflower. South W. Naturalist
28: 231-2306.
‚р, — ue Suzan, W. Hodgson, L. Hernandez
& G. Malda. . Conservation and use of rare plants
by — culo of the U.S./Mexico borderlands.
1: ‚ Oldfield & J. Alcorn (editors), Bio-
aren Conservation and Ecodevelopment.
Westview Press, Boulder, Colorado.
ro Dorothy. Cultural Preservation Office, Southern
Iribe. Ignaci io, Colorado. Telephone interview. 1:
: E
. Dinerstein. 1998. The Global 2000: A
representation approach to conserving the world's most
biologically valuable ecoregions. Conservation Biol. 12:
516—520.
*Pino, Jackson. Middle School and High Sc ue teacher.
Alamo Navajo ош Ala p o, New Mexico. Tele-
phone interview. 15 November
=
*Polacca, Mary. Offic of Hopi Tribal Enrollment. Tele-
phone interview. 10 December 200(
*Poleahla, Anita. Office of Hopi — Preservation.
Telephone interview. 10 December 2000.
Rea, A. M. 1981. California Condor — E A
recovery proposal: Environm. ree 192: €
Ricketts, T. H., E. Dinerstein, Olson & A bs icks.
1999a. Who's where in North fue rica? BioScience 49:
369—381.
Я 3 & 1999}. Terrestrial
Ecoregions of North America: A Conservation Assess-
ment. Island Press, Washington, D.(
*Rollo, Gail. Tribal Administrator. Paiute Tribe of Utah.
Telephone interview. 11 September 2000.
*Salari, Nasim. Program — и Juan Miro
Paiute Tribe. Telephone interview. 11 September 20(
, Merton. C ultura
Tribal Secretary, Tonto Apache Tribe.
ayson, Arizona. Telephone interview. 11 September
JO.
Seaman, P. D. 1993. Users Guide and Index for the A.
ting Collection. Northern Arizona Anthropologic ај
ee 1, F — aff.
*Sec — ro, Tony. President of Tohajilee (С e ito Nava-
j. Telephone interview. 23 October 200(
Volume 89, Number 2
2002
Nabhan et al.
Safeguarding Species
175
Sellars. R. W. 1997. Presen туш Nature in the National
Parks: A Historv. Yale Un y Se s. New Haven.
“Smith, Joni. Acestes ve qoid Camp Verde Y
— Apache Tribe. Telephone interview. 11 Septe “э г
a L.S. Kutner & J. S. Adams. 2000. gre ‘TOUS
Heritage: The Status of Biodiversity in the ed
. Press, Oxford.
*Tohtsoni, Nathan. ° W ho is a Navajo?” Navajo Times. 31
August 2000: 1.
Toledo, У. М. — Biodiversity and к diversity in
Mexico. Different Drummers: Ince
North American Biodiversity 1: 16- 9.
Tuxill. J. & G. P. Nabhan. 1998. Plants and Protected
Areas. World Wildlife Fund чен and People Manual.
Stanle y Thornes Publishing. Cheltenham.
— Office. White нанні Ара he
Arizona. Telephone interview. 13
Also (http://primenet.com) уж ate informa-
уез for Protecting
J
q
Nation.
November
tion).
"Ме mvylewa.
l'elephone interview. 11 S
Wilkinson, €. 1999. Fire on the Plateau: Conflict and En-
durance in the Southwest. Island Press.
Edward. Zuni — nt of Game and Fish.
eptember 2000.
American
Director of Cultural Center.
*Wil Jicarilla
Apache Nation. Dulce, y Mexico. Telephone inter-
view. 13 November 2 o.
E Prescott Indian Tribe (brochure). The Yavapai
Prescott Tw Tribe. 530 E. Merritt; Prescott.
80301.
Arizona
*Web Sites Consulted
(Unless otherwise noted, all web sites listed below were
1)
accessed during December 2000 ог January 200
Ap //www.itcaonline.com) ii c — Hopi.
Kaibab 1 Band of Paiute ен Yava
ДЇ na.com) [Camp Уе PU Pid \pac i |
— yavapai-apache-nation.com) [Camp Verde Ya-
vapai- Apache]
http:// Iced.state. » — [Paiute Tribe of Utah]
1 up н rnment/
е. мт) ‘Southern Ute Tribe]
http: //www.ubtanet.com/~northernute)
tribe |
(http://www.dced.state.ut.us/indian/Today/ute.html)
[Northern Ute tribe]
(http://epa.gov) [Pueblo of Acoma]
(http://www.zuni.k12.nm.us) — Pueblo]
(htup://www.swr-hq.nm.bia.gov/ACRES-19.99.htm) [Bu-
reau of жк Affairs: ا st Region Indian Acre-
As of 31 December 1999” |
http:/ hq.nm.bia.gov/99ABQIf.htm) [Bureau of
dion Affairs:
(http://www.indigenous- ias org) [Indige nous Lan-
titute. Santa Ке, New Mexico: “Endangered
Accessed 5 December 2001 |
i Prescott Tribe|
Li
"n ://
~
[Northern Ute
“Баенов Region Combined Totals” |
guages Ins
Languages Database.”
Richard N. Mack? and
Marianne Erneberg?
THE UNITED STATES
NATURALIZED FLORA:
LARGELY THE PRODUCT OF
DELIBERATE
INTRODUCTIONS!
ABSTRACT
Iver the past 400 years plant immigrants have arrived in the United States in huge numbers, the results of accidental
and deliberate introduction by humans. Very few immigrations, however, result in naturalizations: the environmental
hazards for immigrants in a new range are usually acute and chronic. We traced the history
of introduction among the
naturalized taxa within a group of U.S.
regions and states that span much of the environmental amplitude of the United
States. Despite differences among these regions in physical environments and the length of their histories of plant
ае the — of their naturalized floras that likely arrived through delibe rate introductions is greater than
50 pe t. Many of the first species to become naturalized in the northeastern United States were introduced as food
seasonings or putative sources of medicine. Erstwhile ornamentals are prominent components of all the naturalized
floras, especially in Florida. Some species introduc
‘ed as sources of forage or fiber also became naturalized. Before
1900 many now naturalized species were introduced re peatedly and widely into the United States as seed contaminants
through an extensive international commerce in crop seeds.
The likelihood that the majority of species now naturalized in the United States has a history
cultivation provides a plausible explanation for their
biotic invasion, environmental stochasticity, naturalization, ornamentals, seed catalogs, seed contami-
and post-immigration ¢
Key words:
nants
The mode of entry is unknown for approximately 30 percent.
of deliberate introduction
r persistence in a new range.
Comparisons of the sizes of regional naturalized
floras with the total number of species that have
likely arrived in these new ranges produce a com-
monly shared conclusion: few immigrant species
ever become naturalized (sensu Mack et al., 2000).
Williamson (1993) argued that this difference in the
number of immigrant species, the number that re-
side temporarily (adventive or casual species), and
the number that subsequently become naturalized
differ in turn from each other by an order of mag-
nitude. Accurate determination of these proportions
is rarely possible (but see Holdgate, 1964). Nev-
A
species that arrive in a new range and the number
that eventually persist appears huge (Williamson €
Fitter, 1996). Today more than 30,000 plant taxa
~
ertheless, the discrepancy between the number
are commercially available in Australia (Hibbert,
1999), even more in the United States (Isaacson,
1996); most are routinely cultivated outside glass-
houses or other shelter. Yet even liberal estimates
put the total size of the naturalized flora for each
country at less than 3000 species (Hnatiuk, 1990;
Kartesz, 1994).
Pursuit of explanations for the long odds against
naturalization that confront immigrant species have
long attracted the attention of biologists (de Can-
dolle, 1855
planation would aid substantially the prediction of
; Gray, 1879) because any general ex-
which non-indigenous species will persist. This fas-
cinating topic in ecology and biogeography has re-
cently taken on added importance as nations grap-
ple with erecting legally defensible scientific
! We thank D. R. Gordon, С. Kinter, К. . R.
Wunderlin for valuable — е and олш, and K.
manuscript. Lonsdale, H.
2.
roves
r
Agricultural Library. Support for M. E. was provided by
Effects and. Risks of Biotechnology in — ulture, Theme I
and the Royal Veterinary and Agricultural Univer
? School of Biologic al Sciences
wsu.edu
R ed E Ecology, Royal Veterinary and Agricultural University, 1871 Frederiksberg
address: Depart
— — dk
ANN. Missourt Bor.
SIRO- E (Ca кыы, provided invaluable
advice during the comple tion « б this manusc e "We thank the staf
Academy of Natural Sciences oe the New York Botanic al бш.
Old, K. Schie те — k.
C. Burks and V
Weidema, L. Wester, a
— Il for helpful comments on the
sistance and
the libraries of Washington State | Iniversity, the
the — of Congress, and the National
e Danish Environmental Research Programme (Centre for
II), the Danish National — Research Institute,
‚ Washington wl University, Pullman, Washington 99164, U.S.A. rmackG mail.
C. Denmark. Present
rent of. Terrestrial. Ecology, National Environmental Research Institute, 8600 Silkeborg, Denmark.
GARD. 89: 176-189. 2002.
Volume 89, Number 2
2002
Mack & Erneberg
US Naturalized Flora
177
protocols by which they may exclude harmful non-
1999).
Despite numerous attempts to identify the few that
indigenous species (Glowka € de Klemm,
will become naturalized among a host of immigrant
. Rejmanek € Richardson, 1996; Rei-
1997),
explanation for this numerical discrepancy, much
)
spectes (e.g.
chard & Наша, we still lack a general
less a predictive protocol (Williamson, 1999).
Explanation lies in a better understanding of the
interaction between immigrants and the environ-
ment in a new range. All populations are subjected
to multiple, random environmental events and cir-
cumstances, whether in their native or new range
(Simberloff, 1988).
detrimental events, e.g.. prolonged or acute drought
The consequences of. random
or frost, sudden and massive attack by predators or
virulent parasites, are particularly devastating for
size of a
small populations. Although the sheer
arge population provides some numerical buffering
from repeated losses of a percentage of its mem-
bers. the same percent losses for a small population
can rapidly bring it to extinction, unless members
are replaced through reproduction or immigration.
or both (Menges, 1992, 1998). Many, and perhaps
most, immigrant populations are small founder pop-
ulations for which there is no likelihood that indi-
viduals will be added through recurring immigra-
tion. The randomly occurring losses caused by
ubiquitous environmental stochasticity can readily
prove dire (Mack, 1995).
erant population with attributes that allow it to tol-
Thus, even for an immi-
erate the average expression of conditions in a new
still
chance, repeated occurrence. of extreme environ-
1989: Crawley. 1989).
The raison d'être for cultivation is the deliberate
environment, it. could be destroyed by the
mental events (Crawford.
protection of plant populations from environmental
hazards, including those with stochastic expression.
Such protection can allow the population to reach
a numerical threshold, such that it can sustain loss-
es arising from subsequent stochastic events. M
that threshold size the population may become nat-
uralized. even if cultivation is withdrawn. Thus.
cultivation emerges as a potential counter-force to
environmental stochasticity and may well facilitate
naturalization (Mack. 2000)
If there is a causal link between cultivation and
plant naturalization, we hypothesized that there
should be a corresponding correlation between cul-
tivation and the history of those species that have
become naturalized. For example. in any natural-
ized flora. the majority of species should have a
history that includes cultivation upon entry into the
new range. Clearly, some species have no such link:
immigrants could arrive in a new range and become
naturalized without human dispersal or cultivation
Ridley, 1930; Mack & Lonsdale, 2001
hypothesis is correct, these examples would. how-
(e.g.. ). If our
ever, form a minority of naturalizations.
To test our hypothesis we investigated the mode
of pre-1900 introduction among the naturalized
species within a group of continental U.S. floras
and compared this evidence with similarly con-
structed tallies for naturalized floras from South
Australia (Kloot, 1987), Hawaii (Wester, 1992), and
2000).
substantially
Time to natu-
(Mack et al.
2000). We attempted to circumvent most uncertain
northern Europe (Weidema,
ralization can vary
cases of naturalization by restricting our investi-
galion to species that had arrived in the United
States before 1900.
uralized is based on at least 100 years of residence
in the United States.
As a result, their status as nal-
METHODS
We recognized that our investigation could be
complicated and handicapped by the sometimes
conflicting synonymy of binomial names across the
last two centuries and differences among floras in
the de finition of naturalized species. We selected
published U.S. floras or checklists that identify nat-
uralized species as those that persist without aid of
recurring human cultivation: furthermore, these flo-
ras clearly make a distinction between adventive
(temporary, waif, transient) species and naturalized
species. Our selection criteria are met by recent
floras or checklists for New York (Mitchell, 1986).
Rhode Island (Gould et al., 1998), Florida (Wun-
Чегип. 1998). and North-Central Texas (Diggs et
al.. 1999). We did, however, delete 46 species that
Wunderlin (1998) considered. naturalized because
he reported that each of these species has been
collected only once or has nol been collected poe-
cently. We employed Fernald’s (1950) Gray s Man-
ual of Botany, 8th ed.,
the northeastern quarter of the United States. a re-
because of its coverage of
gion with diverse habitats, and its frequent descrip-
lions on the mode of introduction for non-indige-
nous species. These floras collectively represent
much of the diversity within the continental U.S.
naturalized flora.
We attempted to determine the history of intro-
duction for each naturalized angiosperm in these
floras. We deliberately excluded other plant groups
from our study as the records for some of these
groups (e.g.. eryplogams) are much less complete
and reliable. Information is rarely collected on a
species’ introduction event(s), although some intro-
ductions can be traced detail (e.g... Haughton.
Annals of the
Missouri Botanical Garden
178
Table 1. Summary of the likely mode of entry (deliberate, seed contaminants, as both deliberate & seed contami-
nants, and unknown) for angiosperm species now naturalized in Florida, north-central Texas, the central northeastern
United States, Rhode Island, and New York as derived from published floras (see footnotes 1—5). Numbers in all but
the last column are the percentages of the total naturalized flora that arrived by a mode of introduction.
Delibrate : Deliberate & seed
introductions contaminants contaminants Unknown Number of taxa
Florida! 67 1 l 31 1161
North-central Texas? 62 2 3 34 417
Central northeastern
United States? 59 3 3 5 559
Rhode Island 59 3 3 35 408
New York? 57 3 4 7 391
! Wunderlin (1998).
? Diggs et al. (199€
* Fernald (1950).
' Gould et al. (1998).
5 Mitchell (1986).
Sources for the modes of introduction: Anon. (1882), Austin (1978), Bailey (1906), Barton (1818), Blake (1922), Coon
(1974), Darlington & Wylie (1955). пе (1826). I
(1832), Diggs et al. (1999), Duke (1985),
(1997), Gordon & Thomas (1997), Grieve
— Josselyn (1672), Lamson Scribner ey ls Stre
Mack (1991), Mahler (1980), Mohr (1878). Morton (1976). Morton (1989), Muenscher (1955
Old (1981), үре г (1915), Pursh (1814), Rhoads & Klein (1993
Spencer — Sullé & Maisch (1880), "ке (1919), '
8).
Smith (190
(1969), a rlin (1
1978; Mack, 1991; Schmitz et al., 1991, and ref-
erences therein). In the absence of contemporane-
ous records of a species! introduction, we estab-
lished criteria for identifying the most. plausible
introduction scenario. Species considered here to
have arrived through deliberate introduction have
a history of pre-1900 human use within the species
native range or in the United States. We reasoned
that if a species had been used for centuries as, for
example, a seasoning or herbal in western Europe
(Sturtevant, 1919; Grieve, 1959),
troduced deliberately by human immigrants to the
Dipsacus ful-
it was likely in-
United States (e.g., Nepeta cataria L.,
lonum L., Taraxacum officinale Weber
Crop seeds have been imported to North America
by Europeans for more than 400 years (Viola &
Margolis, 1991). Throughout most of this time seed
cleaning has been either not practiced, ineffective.
or even subverted (see Discussion); consequently,
arriving as a seed contaminant became a likely ac-
cidental mode of entry for some species. Species
were considered to have arrived by this mode if
they had a pre-1850 history as seed contaminants
in Europe or the United States (e.g.. de Schweinitz,
832
For many species a pre-1900 record of use is
in herbals and horticultural compendia
e.g.. Bailey, 1906; Fernald & Kinsey, 1943;
Grieve, 1959; Uphof, 1968), plant materia medicas
recorded
Jarlington (1859), Darwent & Couplanc
— (1989), Fernald (1950), Fernald & Kinsey (1943)
‚ Haughton (1978), Hitchcock (1950), Howell (1959), ‘Hume et al
inge (1977), Long (1922), McCarthy (1888), McCartney (1984),
s Schweinitz
eira et a
(1960),
э). — & Westover Я
, Ridley ei Schery (19€ Schmitz et al. (1991),
Tanaka (1976). U phof (1968), — et al.
—
==
(Sullé & Maisch, 1880), pre-1860 U.S. floras (e.g..
Pursh, 1814; Darlington, 1859), and 19th century
state and federal agricultural publications (see foot-
notes in Table 1). In addition to the primary liter-
ature, we searched 19th century seed catalogs that
had been distributed in the eastern United States
(Mack, 1991,
logs not only report a large array of non-indigenous
and references therein). These cata-
species for sale, but most also state each species’
binomial name and uses (Appendix 1). Fernald
(1950), Wunderlin (1998), and Diggs et al. (1999)
indicated a mode of introduction for some species,
and their determinations supplemented our other
sources.
We initially searched in the pre-1900 literature
for species under their currently accepted names.
Fortunately, many plant species names have been
retained since their initial description [e.g.. Rumex
Vill.]. although we
crispus L., Stellaria media (L.)
usually cannot verify the species’ identification in
a historic account. Some naturalized species in the
modern floras were left without an assigned mode
of introduction because we could not reliably iden-
tify them in the 19th century literature. These were
placed in the Unknown Category. The state or re-
gional floras from which we prepared our tallies all
include some naturalized subspecies and varieties.
We usually could not identify these taxa within the
pre-1900 botanical literature, and unidentified taxa
Volume 89, Number 2
2002
Mack & Erneberg
US Naturalized Flora
179
were also assigned to the Unknown Category. The
Plant (http://
www.ipni.org) was employed as the nomenclatural
International Names Index
standard for plant names and authorities.
RESULTS
Results of our investigation of the history of in-
troduction among these naturalized floras are sum-
Table |
Information on the mode of introduction for
marized in ‚ (Only this summary is provided
here.
each investigated species is available from the au-
thors upon request.) The majority of angiosperm
species within each naturalized flora has a pre-
1900 history of use in their native range or the
United States, or both. The percentage of such de-
liberate use is similar among the floras: lowest with-
in the New York flora (57%), highest in the Florida
flora (07%). Some species have both a history of
use and were 19th century seed contaminants (e.g..
Agrostemma githago L., Bromus secalinus |... Cen-
Convolvulus arvensis L.. Ranun-
culus acris L.) (de Schweinitz, 1832; Fernald. 1950:
1959; Haughton, 1978; Mack, 1991).
species, which form 1 to 4% of the naturalized an-
taurea cyanus |...
Grieve, These
as both
The total
percentage of these naturalized floras with a post-
giosperm floras, are recorded in Table 1
intentional and accidental introductions.
immigration link to cultivation, introduced either
deliberately or as contaminants in crop seeds, is
between 64 and 69%.
under which the remainder (31-37%) arrived
their new U.S.
Similarity
The modes or circumstances
range are unknown.
n the percentages among the floras
surveyed for the central northeast. United States
(Fernald, 1950). Rhode Island (Gould et al., 1998).
and New York (Mitchell, 1986) is due in part to the
similarity of these three species lists. Differences
do occur, however, and may reflect the 40- to 50-
year differences in the collection spans between
Fernald (1950) and the other two accounts. as well
as differences in the intensity of collection. Flori-
da’s climate and less intense pattern of human set-
tlement until the 20th century (Gannon, 1996) has
vielded a naturalized flora that is largely not rep-
the United
States. Most of its naturalized species have sub-
resented elsewhere in conterminous
tropical or tropical native ranges. (Wunderlin.
1998), and many of these species were imported for
potential use as ornamentals and maintained in
government test gardens (Gordon & Thomas, 1997).
This intensity of plant introduction for ornamental
horticulture perhaps explains the high percentage
(69%) of its naturalized flora that has a post-intro-
duction link to cultivation.
DISCUSSION
Our results reveal a strong correspondence be-
tween naturalized species and these species’ delib-
1). Other
species that arrived as seed contaminants in crop
erate introduction and cultivation (Table
seeds would have been the collateral beneficiaries
of cultivation. Together, the proportions in these two
categories support the contention that cultivation,
whether deliberately or inadvertently supplied for
an immigrant species, could have contributed to the
persistence of at least 60 percent of the naturalized
angiosperm species in the regions we assessed.
THE FATE OF
DELIBERATELY INTRODUCED SPECIES
The major agent for spreading plants into new
д oD o
anges around the globe for at least the past 400
vears has been human immigration. In embarking
on an oceanic voyage of colonization, all peoples.
whether Polynesians and Melanesians across the
Pacific (Whistler, 1991)
(Mack. 1999, 2001).
they carry their domesticated plants with them.
Europeans worldwide
have carefully ensured that
This motivation springs from a deep-seated need to
anticipate and resolve the dilemma caused by un-
known. uncertain, or at least not assured sources
of food, fiber, forage, and other essential plant prod-
ucts in a new range. Even if would-be colonists had
advance knowledge that indigenous species in their
new homeland could sustain them, plans for trans-
oceanic. colonization have carefully included the
transport of the germplasm of essential crops in the
initial voyage. Invariably, this transfer of species
that were deemed desirable, if not essential, has
been maintained long after the colony’s survival
1999, 2001).
e rapidity with which the
was assured (Mack,
early colonists 1
eastern North America established European crops
is remarkable: dire necessity is indeed a powerful
1672) cited an
account by Higginson in 1629 in which he raved
stimulus. Tuckerman (see Josselyn,
about the vigor and diversity of the European crops
already available in New England, including beets.
carrots, cabbage. asparagus, radishes, and lettuce.
From the standpoint of species that were to become
members of the naturalized flora, Higginson’s list of
introduced herbs is revealing: sorrel, parsley, cher-
vil, and marigold for pot-herbs, along with sage.
thyme, clary.
anise, fennel, coriander, spearmint,
and pennyroyal as “sweet herbs.”
Cultivated fields and gardens with these Euro-
pean species produced some of the earliest natu-
ralized species in North America. By 1672 Josselyn
reported seeing “Dandelion” (Taraxacum officinale
(Artemisia absinthium L.).
Weber), “Wormwood”
180
Annals of the
Missouri Botanical Garden
and “Black henbane” (Hyoscyamus niger L.) grow-
ing outside cultivation in New England. A century
later, Kalm (1770) found other species, including
Tanacetum vulgare L. and “Datura” (Datura stra-
monium L.), that had also escaped cultivation.
The diversity of species used for medicinal pur-
poses and as seasonings (Grieve, 1959) likely pro-
vided the largest single array of species naturalized
in the United States by ca. 1800. Of the 559 spe-
cies listed by Fernald (1950) as naturalized, at least
65 were in use before 1900 as herbal remedies or
seasonings. Early 19th century seed catalogs often
contained sections devoted to “pot-herbs” and me-
dicinal species: many of the species in these lists
are now naturalized throughout the United States
(Mack, 1991).
list, it is likely that many of these species had been
As suggested in Josselyn's (1672)
introduced much earlier. By the early 19th century
some were already being listed as naturalized in
regional floras, such as Anthemis cotula L., Cynog-
lossum officinale L., Inula helenium L., Nepeta ca-
taria L., Solanum nigrum L., Solidago odora Ait.,
(Barton, 1818).
Pursh (1814) noted the persistence of Cannabis sa-
and Urtica dioica L. In addition,
(“Cannabis sativa”), Cichorium intybus L.
(“Cichorium intybus”), Conium maculatum L. (“Co-
nium maculatum"), Hypericum perforatum L. (“Hy-
pericum perforatum"), Linum usitatissimum L.
tiva L.
(“Linum usitatissimum”), Marrubium vulgare L.
(“Marrubium vulgare”), and Ricinus communis L.
(“Ricinus communis”). Forage grasses, an early per-
ceived deficiency within the native flora, were also
actively imported (Cronon, 1983). As a result, west-
ern European pasture species were members of the
pre-1800 naturalized flora: Aira praecox L., Holcus
lanatus L., Lolium perenne L., Phleum pratense L.,
Poa compressa L., Poa annua L., Poa pratensis L.
(Barton, 1818).
European fruit trees were planted very early in
the settlement of New England (Young, 1846). By
1671, quince, apple, pear, cherry, damson |Prunus
domestica var. insititia (L.) Fiori & Paoletti], plum,
and common barberry were all commonly grown
(Josselyn, 1672).
avium (L.) L., Malus pumila Mill.] have become
Several of these species [Prunus
naturalized but appear innocuous. However, com-
mon barberry’s introduction was soon to plague the
colonists. Berberis vulgaris L. is a host for Puccinia
graminis f. sp. tritici, the stem rust of wheat, and
it would be almost three centuries before common
barberry was controlled effectively in the United
States (Anon., 1937).
Ornamentals, i.e., species introduced purely for
aesthetic reasons, were introduced surprisingly ear-
ly, given the colonists’ need to first establish reli-
able sources of food, fiber, forage, and medicine.
By 1672 Josselyn was commenting on the imported
ornamentals (lavender cotton, hollyhocks, satin, gil-
lyflowers, pinks, English roses, and eglantine) that
he encountered in New England. Eglantine (Rosa
eglanteria L.) is now naturalized in the United
States (Fernald, 1950). Given its early introduction,
it may be one of the first European ornamental spe-
cies to become naturalized in the United States. In
introducing ornamentals, the colonists were greatly
expanding the taxonomic breadth and geographic
range from which naturalized species would be
drawn. We lack adequate records of the market in
ornamental species that emerged in the 18th cen-
tury. Cutler
factly that seeds of an Antirrhinum species were
However, (1785) reported matter-of-
imported b seed-sellers in New England, and a
f "Garden and Grass Seeds, with a
Roots, & Seeds, Just
Imported" was produced as early as 1793 in Rich-
broadside «
choice — tion of Flower
mond, Virginia (National Agricultural Library ar-
chives, as cited in Pennsylvania Horticultural So-
ciety, 1976).
With the apparent proliferation of seed catalogs
by 1800, scores of species were arriving in the
United States from a worldwide list of native rang-
es. Ornamental species (and earliest date of their
appearance in а seed catalog published in the Unit-
ed States) include Lonicera japonica Thunb. (1823),
— camara L. (1804), Melia azedarach L.
(1807), Mesembryanthemum crystallinum L.
Mimosa pudica L. (1804), Myrica faya Dryand
1823). Rhamnus cathartica L. (1807), Rosa mul-
tiflora Thunb. (1826), Schinus terebinthifolius Rad-
„ (1807) (Mack,
1804 a Philadelphia seed merchant
(1807),
—
di (1832), and Ulex europaeus
1991). By
could describe himself as someone who had for
sale, “... South-
Sea Islands, African, and European Seeds, of the
an extensive variety of Asiatic,
most curious and rare kinds: and is daily adding to
his collection, as he avails himself of every oppor-
tunity to procure seeds from all parts of America,
as well as from every part of the world, to whic h
the enterprise of American commerce extends . . .
(B. M’Mahon [1804: 1] “A catalogue of American
Seeds ...”; archives of the National Agricultural
Library, Beltsville, Maryland). The speed with
which regions only recently colonized by Europe-
ans were contributing species to this global traffic
in ornamental plants is impressive. D. & C. Lan-
dreth, seed merchants in Philadelphia, offered nine
Melaleuca species from “New Holland" (Australia)
in their 1832 catalog (archives of the Pennsylvania
Horticultural Society, Philadelphia). In all likeli-
hood, these species had been imported even earlier
Volume 89, Number 2
2002
Mack & Erneberg
US Naturalized Flora
Table 2
Naturalized species in the United States considered by de Schweinitz (1832) to have been deliberately
introduced. Some of the names employed by de Schweinitz are not in current nomenclatural usage, and he did not
include scientific authorities for the species he lis
ted. His names for these species appear parenthetically. The Inter-
national Plant Names Index (http://www.ipni.org) was employed as the nomenclatural standard for plant names and
authorities
/ —— a Roth. (Agrostis p
"stis tenuis Sibth. (Agrostis vulgaris)
Vnthoxanthom мочит L, e 208 odoratum)
чое vulgaris)
‚ Jakoch (Sinapis nigra)
(€ я saliva)
Chelidonium majus L. (Chelidonium majus
Cynoglossum officinale L. (Cynoglossum officinale)
(Daucus carota)
ra stramonium)
Holcus lanatus)
Daucus carota L..
Datura stramonium L. (Datu
Holcus lanatus L. (
Leonurus cardiaca L. (Leonurus cardiaca
.( Marrubium vulgare
а
sativa
Phleum рону nse L. (Phleum pratense)
(Plantago major)
od pratensis
(Rosa —
umex crispu
— А РИ (sic)]
Plantago major L.
Poa pratensis L. (P^
Rosa eglante ria L.
vumex crispus l. (
Rumex pini Ls
Salix alba L. (Salix
—
i ua annuus L. (Scleranthus annuus)
S Stellaria media L. (Stellaria media)
Taraxacum officinale Weber (Leontodon taraxacum)
Trifolium pratense L. (Trifolium pratense)
Trifolium repens L. (Trif lium repens
Verbascum blattaria lL. (Verbascum blattaria)
. (Verbascum thapsus)
(Veronica officinalis)
Verbascum thapsus L
Veronica officinalis L.
to Britain before being introduced in the United
States.
The flurry of published floras that appeared after
1800 provides some of the best evidence we have
of which deliberately introduced species were be-
(“Salix vi-
minalis”) (Pursh, 1814), Crataegus monogyna Jacq.
("C. oxyacantha 1.7) (Barton, 1818),
L.. Salix babylonica 1... and Salix alba L.
coming naturalized. Salix viminalis L.
Acer negundo
(Darling-
ton, 1820) were all recognized as new members of
the eastern U.S. flora. Sometimes the flora’s author
even knew the circumstances of a species’ intro-
duction. Darlington (1826) wryly attributed the es-
tablishment around West Chester, Pennsylvania. of
—
=
Leonurus marrubiastrum L.
Humphrey Marshall, a local horticultural enthusiast
(Wilbert. 1908).
Je Schweinitz (1832) assembled comprehensive
the gardening of
information about deliberately introduced species
that in the United
States. His observations (de Sehweinitz. 1832: 148)
par-
were becoming naturalized
based on major categories of introduction are
ticularly informative: species “purposely brought
hither to be cultivated, for the purposes of agricul-
ture, or for some real or fancied value they possess?
and others that had “been evidently involuntarily
introduced with the imported seeds of agricultural
plants...” АП 31 species that he listed as delib-
erate introductions (Table 2) remain naturalized in
the United States, and a few have become invasive
(e.g.. Poa pratensis l.. Rumex crispus L.. Verbascum
thapsus 1..). Apparently, most of the current worst
invaders in the United States had yet to arrive or
were still undetected [e.g.. Bromus tectorum |...
Lonicera japonica Thunb.. Polygonum cuspidatum
Jal T /
Sieb. & Zucc.. Sorghum halepense (1...) Pers.|.
SEED CONTAMINANTS: EARLY IMMIGRANTS
In an era before herbicides and diligent. seed
sieving and inspection, seed lots of crop species
varied radically in the extent to which they were
contaminated with the seeds of extraneous and un-
wanted species. Some of the earliest records of non-
indigenous plants in North America include spe-
that likely
contaminants, e.g., Rumex acetosella L. and Rumex
1983. and references therein).
Many of these species have been intimately asso-
cies arrived from Europe as seed
acetosa l. (Cronon,
ciated with crops through the strong selection pro-
vided by cultivation and post-harvest storage. Prob-
ably all seed-sown crops have their own array of
seed mimics (Barrett, 1983); each mimic’s phenol-
ogy from germination to seed maturation is under
selection by the cultivation cycle for its associated
crop. Through this close synchrony between the life
eyele of the crop and its mimics, cultivation applied
to the crop could simultaneously benefit the mim-
. leading to their naturalization (Mack. 2000).
We have only a sketchy list of species reputedly
introduced as seed contaminants in the early set-
tlement of the United States (Table 3).
general neglect of seed cleaning, there were. how-
Given the
ever, many possible immigrants among the ruderals
and crop weeds of Europe. The native forage grass-
Ss in New England were deemed so unsuitable as
dnd that by the 1640s a market in European
grass seed had already emerged around Narragan-
Annals
182 of the
Missouri Botanical Garden
Table 3
contaminants among agricultural seeds. Some of the names employed by de Schweinitz are not in current nomenclatural
Non-indigenous species in the United States considered by de Schweinitz (1832) to have arrived as seed
usage, and he did not include scientific authorities for the species he listed. His names for these species appear
parenthetically. The International Plant Names Index (http://www. ipni.org) was employed as the nomenclatural standard
for plant names and authorities.
Achillea millefolium L. (Achillea millefolium)
Agrostemma githago L. ( grostemma githago)
ا ; le
Capsella и (L.) Medik. (Thlaspi bursa
pas storis
Cerastium fontanum Baumg. (Cerastium vulgatum)
Cerastium glomeratum Thuill. (Cerastium viscosum)
Cerastium semidecandrum L. (Cerastium
semidecandrum)
Chenopodiura album L. (Chenopodium album)
Cirsium arevnse (L.) Scop. (Carduus arvensis
Cirsium vulgare may Ten. (Cnicus lanceolatus)
Commelina sativa*
Elytrigia repens var. repens (L.) Desv. (Triticum
repens
Hypericum perforatum L. (Hy peric um perforatum)
Lamium amplexicaule L. (Lamium amplexicaule)
Leucanthemum vulgare Lam. (Chrysanthemum
secu m)
Linaria vulgaris L.
Lithospermum « arvense L. (Litho
olium pere
„ (Plar dies S NONE
innua L. (Poa a a
Polygonum aviculare L. (Polygonum aviculare)
Raphant us — L. (Raphanus raphanistrum)
„) Beauv. (Setaria glauca)
Scop. (Erysimum officinale)
„. (Sonchus oleraceus)
гиса dioica)
„ (Urtica urens)
Veronica — L. (Veronica agrestris)
(Veronica arvensis)
(Antirrhinum linaria
р ermum arvense)
—
>
I
=
=
=
ош
=
У
>
®
a
y
=
=
=
—
SL.
=
D^
=
М
3
=
>
=
~
LIÉ
=
з
>
= (А
د چ
B >
A
~
~
=
=
2
dm
—
—
—
J eronica arvensis L.
* Probably a corruption of Camelina sativa L.
sett Bay (Cronon, 1983, and references therein),
providing ample opportunity for the importation of
(1672:
216) list, “Of such Plants as have sprung up since
seed contaminants. Furthermore, Josselyn’s
the English planted and kept Cattle in New Eng-
and,” includes non-indigenous species that are un-
likely to have been introduced deliberately as pas-
ture often found as seed
species but are
contaminants: “Shepard's purse” [Capsella bursa-
pastoris (L.) Medik.], “Groundsel”
L.], “Sow-thistle” [Sonchus sp.], “Cheek-weed”
[Stellaria media (L.) Vill.]. Cronon (1983: 143) pro-
vided an account from 1652 in which settlers in
[Senecio vulgaris
the New Haven (Connecticut) colony were already
debating without avail as to whether the “
spreading of sorrill [probably Rumex crispus L. or
Rumex acetosella L.| in the corne fields .. .” could
be stemmed. Kalm (1770) commented on European
introductions that he saw along the North American
118) re-
ported the informed opinion of John Bartram and
eastern seaboard in 1748. Kalm (1770:
other American botanists that “Chenopodium album
[Chenopodium album L.| ... is not a native of
America, but has been bruit over amongst other
770: 119) also claimed
that Tanacetum vulgare L., “which grows here and
seeds from Europe." Kalm (1
there in the hedges, on the roads, and near houses,
It likely ar-
rived both through deliberate introduction (Mack,
was produced from European seeds.”
1991) and as a seed contaminant.
Some early 19th century local and regional Unit-
ed States floras also cited species that reputedly
(Crantz.) (false flax).
arrived as seed contaminants. These statements re-
flect informed opinion, rather than documented cas-
es, Nevertheless, they are among the very few ac-
counts of these species in the United States that
are nearly contemporaneous with their arrival. Bar-
ton (1818) described the flora in and around Phil-
adelphia. He listed both Lithospermum arvense L.
and Lithospermum latifolium Michx. as "introduced
among grass seeds from Europe, but now natural-
Pursh (1814) attempted to assemble a flora
of the United States, although most of his own col-
ised.”
lections and exchanges originated in the eastern
states from Virginia northward. He reported that
Anthoxanthum odoratum L. ("Anthoxanthum odor-
atum"), Festuca pratensis Huds. (“Festuca elatior”),
and Centaurea
cyanus (“Centaurea cyanus”)
were either “... introduced with grass seeds from
(Pursh, 1814: 65, 83) or “brought from
Europe with the grain” (Pursh, 1814: 576)
De Schweinitz (1832: 151) provided the first spe-
cific attention to non-indigenous species arriving in
Europe”
the United States as seed contaminants. Although
he did not provide explicit information as to how
he determined which species were “introduced for-
tuitously with agricultural seeds,” his list is none-
theless illuminating (Table 3). These species in-
clude many that remain today as crop seed
contaminants or are ruderals, or both. All are now
naturalized in the United States (Fernald, 1950).
Some in his list also arrived through deliberate in-
troduction, e.g., Hypericum perforatum L. (Haugh-
ton, 1978) and Urtica dioica L. (Uphof, 1968).
Volume 89, Number 2
2002
Mack & Erneberg
US Naturalized Flora
183
SEED CONTAMINANTS: A LONG-TERM MODE OI
IMMIGRATION
The opportunity for non-indigenous species to
arrive in the United States as seed contaminants
grew throughout the 19th century, in part. because
the United States remained surprisingly dependent
on the routine importation of many crop seeds.
Hicks (1895:
dinary list of crop and forage species. including.
"alfalfa. beet,
cauliflower, chicory,
391) maintained that for an extraor-
Brussels sprouts.
kohl-rabi.
ish, salsify. spinach, turnip, the seeds are grown
borage, broccoli.
cress, endive, rad-
abroad, as are also the seeds of many of our grasses,
such as crested dog's tail, sheep fescue. meadow
foxtail, perennial rye grass, and sweet vernal
grass." In addition, “Of the following vegetables
about one-half. of the seeds are imported: Carrot,
eggplant. leek, onion. parsley. parsnip, and pep-
рег.” Large fractions of the seeds needed for do-
mestie production of cabbage. celery, chervil. kale.
and lettuce were also imported (Hicks. 1895: 391).
des (1895) and others (e.g.. Ledoux. 1880:
Ball.
tation had created enormo'is opportunities for the
1898) realized that such massive seed i impor-
inadvertent introduction of unwanted non-indige-
nous species. Furthermore, they recognized that a
cultivated field, carefully tilled to enhance the erop.
was equally advantageous for the emergence of ex-
1880).
had been heightened by recent events.
traneous species (Ledoux. Their concern
An ageres-
sive invader, Salsola kali L.. had arrived in the
1880s as a contaminant in flax seed from Russia.
By 1894 it had already invaded more than 90.000
km? in the wheat-producing regions in the Dakotas
1894). McCarthy (1888) contended that
most of the weed flora in the United States was
(Dewey.
originally introduced and disseminated in the pack-
ages of imported seeds, an unsubstantiated claim
but with some element of justification.
The problem was not, unfortunately, limited to
inadequate seed-cleaning. Foreign seed merchants
deliberately adulterated crop seeds with commer-
: “charlock”
cially worthless species (Sinapis arven-
sis L.) mixed with turnip and rutabaga seeds.
~
[^
“black medic" (Medicago lupulina V.) mixed with
red clover, valu-
able tall fescue and Italian rye (Hicks, 1895: 391).
Ledoux (1880) reported that the seeds of ruderal
English rye mixed with the more
species were routinely gathered in Austria and Ba-
use as seed adulterants. In one 24.5 g
Nobbe (1871) found
3329 extraneous seeds representing 31 taxa. in-
varia for
sample of Phleum pratense 1...
cluding Rumex acetosella L.. Prunella vulgaris L..
Cirsium arvense (L.) Scop.. Sonchus asper (L.) Hill
(“Sonchus asper Villars”), and Spergula arvense L..
In one extreme case, 90 percent of a Canadian seed
lot sold in Michigan as clover consisted of extra-
neous and non-indigenous seeds. The unwanted
seeds averaged 132,000 per kilo in this contami-
nated lot (Hicks, 1895: 393)
ulation that could readily benefit. from any culti-
a large founder pop-
vation upon sowing. Some seed merchants in the
United States were aware of this imported hazard.
The Philadelphia seed firm LV. Faust assured cus-
tomers in its 1868 catalog that “We are most par-
ticular in the purchase of our grass seeds to procure
them from a source where there is no danger of
foreign seeds having become mixed with them, as
we fully appreciate the great damage which some
of these will create if once introduced upon the
soil.”
The response to foreign and domestic contami-
nation was a flurry of state and federal legislation
to examine commercial seeds, including seeds that
had been directly imported from Europe (McCarthy,
1888: Hicks. 1895).
lots from domestic and foreign sources.
Chester (1889) examined seed
Although it
is difficult to distinguish between results for do-
mestic and imported seed lots in his data. the litany
of non-indigenous species he detected is consistent
among all samples: naturalized species typically
found in arable land and roadsides were being re-
(Table 4).
state-appointed seed analysts at the turn of the cen-
peatedly introduced The diligence of the
tury led undoubtedly to curbing the introduction of
United
Unfortunately, these regula-
unwanted non-indigenous species in the
1941).
tory practices were enacted long after many non-
States (Brown.
indigenous species had repeatedly entered the
United States and become naturalized.
Despite the ample opportunity for non-indige-
nous species to arrive as seed contaminants before
1900. we detected few species for which there is a
historic reference to their arrival in that mode: e.g.,
only 14 species within the central northeastern
United States flora.
nevertheless been significant. Some species not re-
This mode of introduction has
ported by any pre-1900 observer likely arrived in
this manner and were simply overlooked. The Unit-
ed States probably derived many of the weeds of
European arable fields simply through the frequen-
cy of their importation as seed contaminants. Fur-
thermore, many species, such as Amaranthus hy-
bridus L., Anthemis cotula V... Capsella bursa-
pastoris (L.) Medik..
cataria, Plantago major L., and Rumex acetosella
Chenopodium album, Nepeta
were continually being introduced and dispersed
through the eastern United States by seed mer-
chants (Table 4). Although these species had ar-
184
Annals of the
Missouri Botanical Garden
Table 4.
in the late 19th ce
Non- sane species detected repeatedly as seed contaminants in domestic and imported crop seeds
гу (Chester, 1889). Some of the names employed by Chester are not in current nonmenc ‘latural
usage, and he did not include scientific authorities for the species he listed. His names for these species appear
parenthetically. The International Plant
for plant names and authorities.
I
Names Index (http://www.ipni.org) was employed as the nomenclatural standard
(Agrostemma githago)
(Amaranthus hybridus)
ithe pr arvensis
Agrostemma githago L.
Amaranthus hybridus L.
Anthemis arvensis
Capsella bursa- -pastoris s(L .) N
Carduus arvensis (L.) Scop. (Cnicus arvensis
n (C the — album)
ulgar
Hypericum perforatu „ (Hypericum perforatu
m)
Leucanthemum ln tu (Chrysanthemum oun
e)
Lithospermum arvense L. (Lithospermum arvens
edik. (C 'apsella bursa-pastoris)
s)
Nepeta cataria L. (Nepeta cataria)
Plantago lanceolata L. (Plantago lanceolata)
Plantago major L. (Plantago major
Polygon ›
Ranunculus sp.
umex acetosella
— ad L. (Rumex crispus)
Setaria
Stellaria media L.
Verbascum thapsus L.
(Stellaria media)
(Verbascum thapsus)
м
rived before 1800, the potential for an increase in
their genetic variation in the United States would
have continued long after these species’ initial in-
troduction, a function of the different European lo-
cales from which later-arriving populations were
drawn (Novak & Mack, 2001).
SPECIES WITH UNKNOWN MODE (
THE UNITED STATES
JF IMMIGRATION TO
We were unable to identify a pre-1900 use or
other mode of introduction for approximately 30
percent of the species now naturalized in the re-
gions we examined. Any assessment of the modes
of introduction in naturalized floras is handicapped
by the paucity and reliability of historic records.
We avoided relying on common names to trace
unless a common
mode of introduction, species’
name has been used consistently for several hun-
=
dred years: e.g., henbane (Hyoscyamus niger L.
foxglove (Digitalis purpurea L.), shepherd's purse
(Capsella bursa-pastoris). This criterion limited our
ability to trace introductions before ca. 1780. Nev-
ertheless, keen observers such as Josselyn (1672)
and Kalm (1770) made invaluable observations.
Additional underestimate of deliberate introduc-
tions was likely because some species were intro-
duced for unrecorded purposes. Their naturalized
descendants are, however, a living link to a pre-
1900 agrarian-based economy in the United States
that relied on few imported commodities. Solutions
to almost all material needs and desires were lit-
erally “home grown." Thus, Hypericum perforatum
was used for medicinal purposes (Darlington, 1859)
and as an object in religious services in 18th cen-
1978).
were used to comb
tury Pennsylvania (Haughton, The dried
heads of Dipsacus fullonum L.
wool (Fernald, 1950). If a plant was deemed valu-
able, its germplasm was imported, even if the like-
lihood of successful cultivation anywhere in the
United States was low 1959; Sullé &
Maisch, 1880). We may never discover all the pur-
poses that our resourceful ancestors had for the
(Grieve,
range of species they so methodically imported.
Species introduced by all accidental (but unde-
tected) modes occur in the Unknown Category, in-
cluding those that can survive attached or within a
vast array of cargo: hay, ballast, packing material,
attached to livestock and clothing (Ridley, 1930;
Mack, in press). But collectively they appear far
less important than deliberate introduction as the
mode by which plants have arrived in new ranges
in the United States in the last 400 years (Mack,
in press).
post- 1900 IMMIGRATION AND NATURALIZATION
Our emphasis has been on the link between pre-
1900 plant introduction and subsequent cultivation
and naturalization. Plant importation has contin-
ued, however, and new species continually become
naturalized. Rejmanek and Randall (1994) reported
that nine species had become naturalized in Cali-
fornia between 1968 and 1993; deliberately intro-
duced species are prominent in this list (Catalpa
bignonioides Walt., Nerium oleander L., Pinus pinea
L., Pinus halepensis Mill.). The post-1900 growth
of the naturalized flora in the United States has
likely been substantial. Henry and Scott (1981:
318) tallied the dates of introduction for the “alien
component of the spontaneous Illinois vascular flo-
»
ra," species that apparently include all naturalized
but also adventive species as well. They concluded
that the woody and herbaceous component of this
non-indigenous flora before 1922 was composed of
440 species; 163 species were added between 1922
and 1955, and another 208 non-indigenous species
were added from 1956 to 1978. Many of these post-
Volume 89, Number 2
2002
Mack & Erneberg 185
US Naturalized Flora
Table 5.
naturalized angiosperm floras of S
outh Australia, Hawaii. and Northern Europe. Numbers in
Tallies for the likely mode of entry (deliberate, accidental, deliberate & accidental, and sê for the
each column except the
last are percentages of the total naturalized flora surve yed. Sources for the modes of introduction are listed in the
footnotes.
Deliberate & | Unknown mode Number
Deliberate Accidental accidental of introduction of taxa
South Australia! STI 24 0 19 901
Hawaii? 51 39 0 1 813
Nordic continental countries” \ 36 14 13 <8 559
North Atlantic Islands” " 16 ol 9 14 10
! Kloot ( 1981).
? Wester (1992).
' Weidema (2000).
VN
(2000)
from the data of Weidema (2000).
1900 plant immigrants to Ilinois would have been
if not all.
Hamilton,
including most,
(Reichard &
deliberately introduced,
the woody
1997)
The proportions among deliberately and acciden-
immigrants
tally introduced 20th century immigrants that have
become naturalized are better documented in Aus-
Of the 290 weed,
cursions into Australia from 1971 to 1995 that have
tralia. i.e.. deleterious plant. in-
led to naturalizations, 65 percent of these species
were introduced as ornamentals, and an additional
7 percent arrived as intended additions to agricul-
1998).
these species have had some degree of post-immi-
Groves,
—
ture
gration cultivation. The proportion of species arriv-
ing in Australia through deliberate action continues
pattern set into motion centuries earlier among
European colonies and their trading partners.
Detailed examination of the fate of introduced
woody ornamental species across much of the 20th
2001),
suggests an additional aspect of the importance of
century in. Canberra, Australia (Mulvaney,
post-introduction cultivation. Mulvaney (2001) con-
tended that the probability of a species becoming
naturalized is a direct function of the number of
1909
A similar explanation has
—
recorded plantings of from
through the mid 1980s.
been proposed to account for the naturalization of
the spec les
non-indigenous birds in New Zealand: persistence
correlates with the intensity of the introduction ef-
Veltman et al., 1996). The more separate op-
—
forts
portunities for non-indigenous species to be culti-
vated, the greater the probability some of its
immigrants will be initially spared the full force of
environmental stochasticity in the new range.
! Numbers are means of percentages for Greenland. Iceland. the
Thus, more than two-thirds of
umbers are means of percentages for Norway, Sweden. Finland. and Denmark as derived from the data of Weidema
Aland Islands, and the Faeroe Islands as derived
NATURALIZED SPECIES ARISE FROM DELIBERATE
INTRODUCTIONS: A RECURRING PHENOMENON
WORLDWIDE
Our chief goal was to evaluate in an objective
manner the hypothesis that much of the U.S. nat-
uralized flora has a historic link in its mode of entry
to deliberate: introduction. and post-immigration
cultivation. At the
benchmark for rejecting this hypothesis would be
outset, we decided that our
the failure to detect that even half of the naturalized
species had a history of pre- 1900 use. Our hypoth-
esis appears supported by our tallies (Table 1).
The plausibility of this link is further supported
by evidence gathered among naturalized angio-
Wester (1992) examined
the modes of introduction within the Hawaiian nat-
кф,
sperm floras worldwide.
uralized flora (Table 5). He concluded that the ma-
jor mode of introduction had been deliberate
and that a large fraction of these species
had been introduced in ornamental horticulture. By
coincidence, Kloot (1987) also found that at least
57 percent of the South Australia naturalized flora
owed ils arrival to deliberate transport. (Table 5)
Esler (1987) determined the modes of introduction
(imported for use in either horticulture [including
i shelter tree
timber and
specie r agriculture
| crop. pasture, and land —— or accidental)
for the 303 angiosperm species now naturalized in
urban Auckland, New Zealand. Almost 93 percent
of this naturalized flora was introduced deliberately,
although the degree of post-immigration cultivation
probably varies.
the mode of introduction has been as-
Denmark.
Norway. and Sweden along with
Recently,
sessed for non-indigenous species in
Finland. Iceland.
186
Annals of the
Missouri Botanical Garden
the Aland Islands, the Faeroe Islands, and Green-
land (Weidema, 2000). Here again, many natural-
ized species appear to have a history of deliberate
introduction (e.g.. /nula helenium L., Lychnis chal-
cedonica L., Ornithogalum nutans L., Syringa vul-
garis L., Spiraea salicifolia L.), but the role of ac-
cidental introduction appears to be much larger
than in the United States South Australia.
Among Nordic countries in continental Europe.
about one-third of the naturalized species are con-
sidered to have been deliberately introduced, but
more (44%) on average are believed to have been
introduced accidentally as seed contaminants of
cargo or carried by domesticated animals (Table 5).
These values from northern Europe require fur-
Among Nordic investigators
or
ther interpretation.
there is apparently neither a consensus on the def-
inition of “naturalized” (the values reported may
also include adventive species for the floras of some
~
countries) nor on an arrival date before which an
immigrant species is deemed native (Weidema,
2000). Potentially more important is the much lon-
ger history of agriculture in northern Europe than
farming by European colonists in the United States
or South Australia. Several millennia of raising
crops in northern Europe has given ample oppor-
tunity for species to have been introduced, both
deliberately and inadvertently (Iversen, 1973). For
some of these species, their mode of introduction
is unknown [e.g.. Helleborus foetidus L., Potentilla
micrantha Ramond ex DC.. Digitalis lutea L., Sil-
ene tartarica Pers.| (Weidema, 2000). Furthermore,
some fraction of those species now considered na-
tive arrived so long ago with human settlement that
any identification of erstwhile deliberate use is
problematical. The much shorter histories of plant
introductions into the “New Europes” in North
America, Australia, and South America provide us
with a sharper picture of the causes of plant natu-
ralization than can be reconstructed from the re-
cords of plant dispersal by humans in Europe.
Needed now is experimentation that bridges the
gap between two growing bodies of information:
knowledge of the modes of plant introductions and
naturalizations since A.D. 1500 (Kloot, 1987; Rei-
chard & Hamilton, 1997; 1998; Mulvaney,
2001) and understanding of the stochastic forces to
which immigrant populations are usually vulnera-
ble (Menges, 1992, 1998; Mack, 1995, 2000). The
design of experiments on the fate of immigrant pop-
ulations (Panetta & Randall, 1994) could benefit
from clues derived from the history of plant intro-
ductions. Experimental variables, including the siz-
es of the immigrant populations, the initial entry
locales in the new range, and the character and
Groves,
extent of cultivation, could duplicate what is known
about a species’ early history in its new range. Once
such investigations become widespread, our under-
standing of the causes of plant naturalization and
subsequent invasion will likely become much
clearer (Williamson, 1999; Williamson & Fitter,
1996).
Literature Cited
1882. Sample No. 1,398. North Carolina Agric.
. Sta. Annual Rep. p. 147.
1937. A Report on the I — ~ in Bar-
— Ersdia ation. During the Ca 1936 in-
cluding Summarized — or the Period 1918-19: 30.
U.S.D.A. dE Pl. Dis. Ctrl
Austin, D. 1978. Exotic plants and their effects in
ssi does Environm. Conservation 5: 25-34,
Bailey, L. Юю. Cyclopedia of American Horticulture,
Comprising (ipee n for Cultivation of Horticultural
Plants, Descriptions of the 'ruits, Vegeta-
bles, Flowers and Ornamental Plants Sold in the United
States and Canada, Together with Geographical and
Biographical Sketches. 4th ed., 6 vols. Macmillan, New
—
Ex
Species of
ОГК.
Ball, С. К. 1898. Seed testing: Its importance, — and
some results, with a partial bibliography. lowa State
Coll. Agric. Mech. Arts, Contr. from the Botanic al De-
partment,
Barrett, S. С. н. 1983. Crop mimicry in weeds. Econ. Bot.
37: — 32
Barton, W. P. C. 1818. Compendium Florae Philadelphi-
cae: — a Description of the Indigenous and
Naturalized Plants found within a Cire 2 Е Ten Miles
around Philadelphia, Vols. 1 and 2. M. Carey & Sons,
^
M
1922. Two new western weeds. Science 55:
— A brief record of seed testing in the Unit-
es Department of Agriculture. Pp. 32-39 in M
T Munn (editor), History of the Association of Official
Analysts. Assoc iation of Official Seed Analysts,
iar. A. de. 1855. — ا е вЫ
e. des Faits Principa des Lois Ca
ant la ирне ean ane des Plantes ds
Г Epoque x uelle. 2 vols. V. Par
— F. 1889. s ГЕ De bane Coll.
Exp. Sta. Ball 5: 5-
zT
Masson,
Agric.
Coon, N 1974. The o of Useful Plants. Rodale
Press/Book — Emmaus, Pennsylvania.
Crawforc . M. 1986 Sudie 's in Plant Survival: Eco-
logic al Са азе "Histories of Plant Adaptation to Adversity.
Blackwell Scientific, Oxford.
Crawley, M. J. 1989 Chane e and timing in biological in-
vasions. Pp. 407-423 ¡ Jrake et al. (editors), id
logic ‘al Invasions: A Global Perspective. Wiley,
ork.
— W. 1983. Changes in the Land: Indians, Colo-
nists, and the Ecology of New England. Hill and Wang,
New Y
Cutler, М. 1785. An account of some of the vegetable
Dia шен, мна growing in this part of America.
Amer. s Sci. Mem. 1: 396—493.
Darlington, C. p. & A. P. Wylie. 1955. Chromosome Atlas
of Flowering Plants, 2nd ed. Allen & Unwin, London.
ОГК.
Асаа.
Volume 89, Number 2
Mack & Erneberg 187
US Naturalized Flora
Darlington, W. 1826. Florula Cestrica: an Essay towards
a Catalogue of the Pheenogamous Plants, Native and
— м Growing in the Vicinity of the Borough of
»st-Chester, in C * 'sler County, Pennsylvania. 5. Sieg-
friec *
1859, American Weeds and Useful Plants, 2nd
ed. (revised by G. Thurber
Darwent, A. T. bu uf:
Оор panic wate Weeds 14: : 31: 31
Dewey, L. H. 189 e Russian thistle: [ts (m as a
weed in the U р States, with an ace — of the means
— for its eradication. U.S.D.A. Div. Bot. Bull.
ES
Diggs. G. M.. Lipscomb & R. J. O'Kennon. 1999,
Shinners & г. s Illustrated Flora of North Central
x xas. Botanical Research Institute of Texas. Fort
d | 1985. CRC Handbook of Medicinal Herbs.
CRC AUN жа, ‘a Raton.
Esler, A. К. 1987. The м: of plants in urban
Auckland Zealand. Catalogue
of naturalised
s. New Zealand J. ES 25: 53 8.
989. Commelina caroliniana E ommelina-
eae): À вине сб species in the United States is
an old — ‘tion from Asia. Taxon 38: "tan
Femald. L. 1950. Gray’s Manual of Botany. Sth ed.
sana Press, Portland. Ore
& A.C. Kinsey. 1943. Edible Wild Plants of East-
Idlewild Press, Cornwall-on-Hud-
—55€
ern. North — “a.
son, New
Ferreira, J. F. s. 1. Simon & J. Janick. 1997. Artemisia
annua: Botany, Шейше, pharmacology. Hort.
19: 319-371.
Gannon, M. (editor). 1996. The New History of Florida.
Univ. Press — Gainesville.
Glowka, I e Klemm. 1999, International instru-
ments, proc organizations and non-indigenous
species introductions: Is a — to the Convention
on Biological Diversity necessary? Pp. 389-405 in O
T. Sandlund, P. J. Schei & A. Vike n adiu Invasive
p and Biodiversity Management. Kluwer, Dor-
drec
ddon. 5. R. & K. 1997.
y de ig п ry, screening and ге
ps Pp. 21-37 in D. Simberloff, D. C. Schmitz & T
. Brown (editors), Strangers in Paradise. Island Press,
C
esses,
Thomas. Florida's invasion
E sy L. L., R. W. Enser. R. E. Champlin & I. H. S
998. Vascular Flora of Rhode Island. Rhode Island
Natural History Survey, Kingston, Rhode Island.
Grey, А. 1879. The pertinacity — tque of
— a Sci. Arts 18: 161—
Grieve, M. 1959. A го 1 Herbal, D и C. F. Leyel
(editor). ane. 3 Yor!
; . Recent incursions of weeds to pos
Tech. S
Groves, R.
tralia. nC à — л Systems,
No. 3. University of Adelaide, Ade laide.
Haughton, C. S. 1978. Green Immigrants: Plants that
Transformed America. Harcourt, Brace and Jovanovich.
New York.
Henry, R. . R. Scott. 1981. Time of introduction
of the кж шол of the — ous Illinois vas-
cular QE — . Midl. Naturalist 106: 318—324.
Hibbert, M. The Aussie Plant Finder 1999/2000.
Horum po New South Wales.
Hicks, G. H. 1895. Pure seed investigations. Pp. 389-408
g7
in Yearbook of the United States Department of Agri-
Е D.C.
50. Manual of the — the Unit-
| — 2nd a (rev. by A. Chase). U.S.D.A. Misc.
00. Washington. D iC.
Hnatiuk, as J. (editor). 1990. Census of Australian кс
cular Plants. "i Flora and Fauna series: no.
Australian Pt Pub. Service, Canberra.
Holdgate. M. W. n experimental introduction of
94. A
p to the pie ы Bull. Brit. Antarc. Surv. Bull.
13-1
— J. T. 1959. V وف ar data on weedy thistles
in western өп America. Le . Bot. : 17-32.
Hume, L.. J. The biology of
Canadian weeds. 60. ppm — ulus L.
J. Pl. Sci. 63: 959—
Isaacson, R. T. 1996. S ca List of Plants and Seeds, 4th
ed. Andersen Horticultural Library. University of Min-
nesota — Minneapolis.
Iversen, J. 1973. The ica of Denmark’s nature
since the last е ial. (Translation from Danmarks natur.
1: 345-445, 1967, by M. Robson.) C. A. Reitzels For-
lag. a ques n.
Joselyn, 2. New-England’s Rarities, Discovered in
Birds, Beasts, Fishes, Serpents. and Plants of that
Country. [Reprint. with introduction by E. '
105-238 in Transactions and ( ‘ollections of the
American Antiquarian Society, Cambridge, 1860. |
Kalm, P. 1770. The America of 17 50; Peter каз travels
i ion of 1770.
Martinez & K. E
бырай,
94. A =r Mie Checklist of the Vas-
dila Flaca Bi the United States, Canada. and Green-
land, 2 08: a ed. Timber Press, Portland, Oregon.
api Р. . The naturalised flora of South Australia
ts origin, S d — — forms
i signifi ance. J. Adelaide Bot. Gard. 99-111
„amson Scribner, ЕЁ 1809. Weeds he Mas ron
Office, Augusta, Maine.
pus A. R. 1880. Seed examinations.
ric. Exp. Sta. Annual Rep., pp. 22—
TE B. 1922. Sonchus uliginosus occurring in the Phil-
adelphia area. Torreya 22: 91-98.
McCarthy. G. 1888. The purity and vitality of seeds, with
recent tests of grass and clover seed sold in North pas ar-
olina. Коп т — Exp. Sta. Bull. 59: 4—32
—
ous
Carolina
л
McCartney, С. rehids by ba side of the ded
Amer. = bae Ed. 53: 18
Mack, R. The aa ы trade: An early
tem " Me Econ. Bot. 45: 257-273
— ——-. 1995. Understanding the processes of weed in-
vasions: The influence of environmental stoc bass
Pp. 65—74 in C. Stirton (editor). Weeds in a Changing
World, Symposium
British Crop Protection Council,
I pira No. 64. Brighton.
— ———. 1999, The motivation for poring
invasive plant species: A primal urge? Pp. —502 in
ите of "n VI ао Rangeland Con-
gress, Townsville, Australia.
acie Al, 00. Lim à fosters plant naturalization by
reducing environmental stochasticity. Biological Inva-
sions 2(2): 111—122.
— ———. 2001. Motivations and consequences ^e the hu-
man a renal of plants. Pp. 23-34 in J. McNeely
(editor), The Great Reshuffling: Human — nsions in
Invasive Alien Species. International Union for
—
188
Annals of the
Missouri Botanical Garden
Conservation of Nature, Gland, Switzerland, and Cam-
bridge, U.K
—. In press. Global plant — sal: Pathway
modes and circumstances. /n G. Ruiz & J. Carlton (ed
itors), Global Pathways of Biotie pcre Island
Press, Washington, D.C.
— ——— & W. M. Lonsdale. 2001. Humans as global е
dispersers: Getting more than we bargained for. Bio-
Science 51: 95-102.
——, D. Pan W. M. Lonsdale, H. Evans, M.
Clout € F. A. Bazzaz. 2000. Biotic invasions: Caus
epide mol Vou consequences and control. Ecol.
Appl. 10: 689—710.
Mahler, M. 3 1090, Limnophila—A new exotic pest.
в 2: 4-
Menges, E. S. 1992. Stochastic modeling of extinction in
plant populations. Pp. 255-275 in P. L. Fiedler & 5
Jain (editors), Conservation Biology. Chapman and
Hall, New York.
1998. Population viability analyses in —
че and opportunities. Trends Ecol. Evol.
Mite he ^a R. S. (editor). 1986. A Checklist of New York
State Plants. Unive a of the State of New York, State
Education "i pt., Alb:
Mohr, €. 187 —— piant introduced into the Gulf
States. Bot. p 3: 42-10.
Morton, J. F. 1976. Pestiferous spread of many ornamental
and fruit Ае! ies in South Florida. Proc. Fla. State Hort.
Soc. 89: :
989. C ree ping Indigo (Indigofera spicata Forsk.)
Flo
— rd to herbivores in Florida. Econ.
Bot. 43:
Muensc her T, E P „ 1955. Weeds, 2nd ed. Macmillan,
New Yor
Mulvaney, M. 2001. The effect of introduction pressure on
the naturalization of ornamental woody plants in south-
eastern Australia. Pp. 186-193 in R. H. Grov
Panetta, J. G. Virtue (editors), Weed Risk ppa *
CSIRO, Melbourne.
Nobbe, F. 1871. Mittheilungen aus der physiologischen
versuchs-station Tharand. IV. Botanische untersuchun-
gen kauflicher landwirthsc haftlic her Samereien. Land-
wirtschaftlichen Ve poe ‘hs-Stationen bd. 13: 233-235.
Novak, S. J. € R. N. Mack. 2001. Tracing plant introduc-
tion and spread into naturalized ranges: Genetic evi-
депо e from ен tectorum (cheatgrass). BioScience
Oakley, R. A. 4 H. L. Ау slover. 1910. — rcial vari-
eties of alfalfa. U.S.D.A. Farmers Bull. 1-23.
Old. R. R. 1981. iin skeletonweed (С нін d ea
L.): Its Biology, Ecology and Agronomic Hist
Thesis, ЕИ State University, Pullman
Panetta, E D. € R. P. Randall. 1994. An assessment of
— ability of Emex — Austral. J. Ecol.
19: 76—82.
Pennsylvania Horticultural Society. 1976. From Seed to
Flower: Philadelphia, 1681-1876: A Pun ultural Point
of View. Pennsylvania Horticultural Society. Philadel-
phia.
Piper, C. V. 1915. Forage Plants and their Culture. Mac-
— New York.
Pursh, F. 1814. Flora Americae Septentrionalis. White.
C e London.
Reic Бай, S. H. & С. W. Hamilton, 1997. Predicting in-
vasions of woody xt introduced into North America.
)
Conservation Biol. 11: 193-2
Rejmanek, М. & J. M. Randall. 1994. Invasive alien
plants in California: 1993 summary and aeter
with other areas in North America. Madrofio 41: 161—
177.
). M. Richardson. 1996. What attributes —
some * species more invasive? Ecology 77: 165
61.
— A. F. ‚ M. Klein. 1993. The Vascular Flora
of dud шша] Checklist and Atlas. Amer-
ican P — q Society, Philadelphia
Ridley, H. The Dispe rsal of Plants Throughout
the Wor a | ere Ashford, Kent.
Sche ery, R. ы, . The migration of a plant. Nat. Hist.
: 41—
Se tu D: ‚В. V. Nelson, L. E. Nall & J. D. Schardt.
99]. Exotic aquatic plants in Florida: A historical per-
spective and review of the present aquatic plant regu-
ation program. In T. p. Center et al. (edit ors), Proceed-
ings of the Symposium on Aquatic Pest Plants.
Technical report NPS/NRE VER/NRTR-91/06. U.S. De-
partment of Pu eres National Park Service, Denver.
Schweinitz, L. D. de. ‚ Remarks on the plants of Eu-
rope which en — naturalized in a more or жа
degree, in the Unit E States. Ann. Lyceum Nat. His
New York 3 8-15
Simberloff, D. 1988. The ат of population and
ое — to conservation science. Annual
col. S 19: 473-5 "
E J. G. 1900. diga and Forage Plants. U.S.D.A.
Div. Agrostol., (p rev. ed.).
Spencer, N. R. 1984. Ve ‘lvetleaf, Abutilon theophrasti
cla eae), histon — economic impact in the United
States. Econ. Bot. 38: * *
Stillé, A. & J.M. М b The National Dispensa-
tory. Containing the Nara s y, Chemistry, Phar-
macy, Actions and Uses of Medicines, inc luding those
Recognized in the — opoeias of the United States,
Great Britain, and Germany, with Numerous References
to the French E _ x, H. C. Lea, Philadelphia.
Strange, Ң. le. A History of Herbal Plants. Arco,
New York.
Sturtevant, E. L. 1919. Sturtevant’s Edible Plants of the
World. /n U. P. Hedrick (editor), New York Department
of Agriculture and Markets. Annual Report; 27th,
1918-19, v. 2, pt.
Tanaka, T. 1976. 1 Tanaka's Cyclopedia of Edible Plants of
the World, S. Nakao и Yugaku-sha; distributed
by Ke igaku Pub., То
Uphof, J. €. T. 1968. Dictionary of Economic Plants. Wel-
don & We се js w Yor
Veltman, C. J.. ee & M. J. Crawley. 1996. Correlates
of — success in exotic New Zealand birds.
Amer. Naturalist 147: 542—557.
Viola, H. J. € С. М: irgolis (editors). 1991. Seeds of
Change: A Quintcentennial Comme moration. Smithson-
ian Institution Press, Washington, D.
Weidema, I. (editor). ped — ‘ed Snes ies in the Nor-
dic е — 213, Axe
Weldon, L. W.. uc ios & D. S. Harrison. 1969.
Common Minim weeds. Pp. 1-43 in USDA. Agricul-
tural Жү No. E
Wester, L. 1992. Origin and distribution of adve BUE alien
quas СЕ іп Hawai’ ii. Pp. 99-154 in С. P. Stone
nith & J. Т. Tunison (editors), Alien is In-
vasions in Native ben ms of Hawaiʻi: Management
and Research. University of Hawaii Cooperative Na-
tional Park Resources Studies Unit, Honolulu.
“=
=
=
T
Volume 89, Number 2
2002
Mack & Erneberg 189
US Naturalized Flora
Whistler, W. A. 1991. Polynesian plant introductions. Pp.
11-66 in A. С. Cox & S. A. Banack (editors). Islands,
Plants, and Polynesians: An Introduction to Polynesian
‘age i Dioscorides, Portland, Oregon.
Wilbert, M. I. 1908. Some early botanical and herb gar-
r J. — 80: 412-427.
Invaders, weeds
FE
dens re
Williamson м. ; and the risk from
etically modified organisms. Experientia 49: 219—
. 1999, ои ا 22: 5-12,
— А. — 199 s
le varying success of invad-
ok ор) : 166 * 660.
— RI 1998, S to the Vascular Plants of
"lorida. Univ. Press Fk rida, — le.
1846. Chronic les of the Pilgrim Fathers of wa
Colony of Plymouth, from a to 1625: Мө ow First Ce
lected from Original Records and С
Printed Documents. and имне d with Notes "s. L ittle &
Bostor
Young, A.
Brown,
APPENDIX |
Pre-1900 use of some of the perae angiosperm
— that — the tallies in Table | was indicated
by their sale i > 19th e (Archives
of the Rational rae cultural | ibrary. Beltsville, Maryland).
entury seed —
Allen’s annual catalogue of seeds. 1870. New York. New
\ze 1 Вані h. 1854.
seeds for sale. Boston,
Barr. G. H. and Co. 1853. Catalogue of agricultural and
\ descriptive catalogue of flower
Massachusetts.
— ultural implements field and garden seeds and
fertilizers.
J. clone and Son. 1807.
and herbaceous plants, indigenous to the United States
p America. Philadelphia, Pe nnsylvania
1870.
A catalogue of trees. shrubs.
Bliss and Sons. Catalogue al seeds. Spring-
|
Pe | i eng husetts
Charles H. B. Breck.
tree, ee and grass seeds, for sale
store. Boston, Massachusetts.
Robert Carr. 1828 Bartram’s catalogue of trees. shrubs.
and herbaceous plants, indigenous to the United States
of America. Philadel Ма, Pe nnsylvania.
Bernard M'Mahon. 1804. catalogue of American seeds,
& c. Philadelphia, воа
Bernard M'Mahon. s. d. A catalogue of
flowers, tree, shrub and
Pennsylvania
Pomeroy and Marshall. 1857.
1842. Catalogue of vegetable, herb,
al the New Seed
garden, herbs.
grass seeds. Philadelphia.
Sar title), Mobile. Alabama.
Warren's Garden and Nurseries. 1845. Annual descriptive
catalogue of fruit and ornamental trees. Boston. Mas-
sachusetts
Me essrs. Winship. 1833. Annual catalog of the
ornamental trees and plants, cultivated for sale at the
Brighton Nurseries by Messrs. Winship. Boston, Mas-
sachusells.
Winter * Co. Catalogue of fruit and ornamental
— innéan QU y Dude n and nursery (late of the
messrs. Princes).| Flushing. New Yor
Winter and Co. 18 14-1845. Catalogue at fruit and orna-
mental trees. [Linnean Коше
(late of the messrs. Princes).] Flushing, New Yor
fruit and
Garden and nursery
THE DODO WENT EXTINCT
(AND OTHER ECOLOGICAL
MYTHS)
Stuart L. Pimm!
ABSTRACT
The scientific consensus is that human impacts are driving species to extinction hundreds to thousands of times
—
they deserve, | examine four concerns.
aster than ed from the natural background rate. Crities challenge this. Perhaps giving them more credit than
First, that the extinction crisis is not real. It is and high rates of extinction are
the rule, not the exception, within well-known taxa. The second criticism — the problem as one restricted just
to islands. It is not. Island species have special аа н but they
ranges than are continental species with the same range
species with small a ranges that are threatened by "nn impacts.
the clearing of forests from eastern
that became extinct followin
e far more locally abundant within their
There are de numbers of loca lly rare, continental
third criticism notes the few
species
North America in the 19th century,
casting doubt
upon the re PERRA be — habitat loss and spec ies loss. — of this case history shows that exactly as many
spec ies 0 MI
such as =a eae Asia and the
This leads to the final criticism: that there have been too
deforestation is more recent and species do not go extinct — Some doomed species can linger for decades
as did the now-extinct species in eastern North Ame ric
Key words:
were lost as expected, for the region had v
ry few species to lose. Extensions to species-rich areas
Atlantic. coast. of Brazil conim the expected calibrations with an interesting caveat.
Forest losses predict the number of threatened species—tho
n the verge of extinction—not the number of extinctions.
ew recent extinctions. The re ply is that in these regions the
deforestation, extinction, species-area curves.
Among scientists there is a broad consensus that
species are going extinct in unusual numbers. I will
not assemble the evidence for this directly because
there are recent reviews (May et al., 1995; Pimm,
2001; 1995). Rather, I wish to tackle
the critics who dispute this consensus. Whatever
Pimm et al.,
one thinks of them and those who finance some of
them, however one scorns their willingness to ig-
nore volumes of inconvenient facts, the critics per-
sist. They will likely continue to do so while indi-
viduals gain financially
environmental destruction.
from short-term
Over the last decade, |
have listened to these critics and, perhaps giving
them more credit than they deserve, assembled the
science to rebut them directly. The synthesis I pre-
sent here is one based largely on my own work on
birds. This is not because it is unique—far from it;
there is an abundance of evidence to counter these
critics. Rather, it is an attempt to lay out cohesive,
linked arguments into a recipe that readers can
readily apply to other taxa.
There are four criticisms.
The extinction crisis is not real. Rather, it is
(Budiansky, 1993). It is the
"facts, not the species" that are endangered (Simon
& Wildavsky, 1993: A23), the estimates of extinc-
“strident, inconsistent, and data-free"
a “doomsday myth”
tion rates are
(Mann & Plummer, 1995). Since the humid tropical
forests of the Amazon, Congo and New Guinea, and
elsewhere hold the majority of species, their fate is
closely tied to the fate of species. Stott (1999) had
this to say about them: “Tropical rain forest’ does
not exist and never has existed.”
Those who accept that unusual numbers of
extinctions have occurred can still proceed to dis-
miss their significance. “The dodo went extinct”
proclaims the Oxford English Dictionary. Most re-
cent like the dodo, have been on is-
The implication is that island species are
wimpy, naive, and unsophisticated. Perhaps island
species had it coming to them and the urbane, so-
phisticated species that populate continents may
not share their fate.
extinctions,
lands.
3. Habitat destruction does not cause extinc-
tions—look at eastern North America, Budiansky
(1993) urged. Projections of future high extinction
rates such as those by Wilson (1988) and Raven
(1988) combine well-documented rates of tropical
forest destruction and a model to predict species
loss from habitat loss. How good are those predic-
tions? Eastern North America was cleared of its
deciduous forests from 1750 to 1900, vet suffered
few known extinctions. Critics argue that we sim-
ply do not know how to predict the numbers of
! Center for Environmental Research and Conservation,
York, New York 100027, U.S.A. StuartPimm@aol.com
ANN. Missourt Bor.
MC 5556, Columbia University, 1200 Amsterdam Ave.,
New
GARD. 89: 190-198. 2002.
Volume 89, Number 2
2002
Pimm
Species Extinction
species that will be lost as tropical forests disap-
pear.
4. In what is mostly a rehashing of earlier myths,
Lomborg (2001) seemed to be asking where are the
bodies to prove an extinction crisis? Some early
efforts did indeed suggest that there should be lots
of extinct species by now. For example, “one sev-
enth to one fifth of all species” extinet within what
would now have been the last two decades (Barney,
1980: 328). There are not nearly enough, though to
continue the metaphor, there are the requisite num-
ber of seriously wounded ones. So are these species
really dying off at the expected rate—or are our
concerns about them misplaced?
l consider each of these myths in turn.
1. The EXTINCTION CRISIS Is Nor REAL.
Has humanity increased extinction rates beyond
the background rates expected without our im-
pacts? Those who argue that we have not are claim-
ing that far too few species have gone extinct in the
recent past. Where should we look for the extinct
species that would reject this assertion?
Pacific islands are the obvious place to start, for
they were the planets last habitable areas to be
colonized. Polynesians reached them only within
the last 1000 to 4000 years. The evidence of human
impact is freshest here. (The evidence of human-
caused mass extinctions in Australia, Madagascar,
and the Americas grows more compelling each
1999). Pacific
birds provide unambiguous evidence of massive ex-
tinction (Pimm et al., 1994; Steadman, 1995). The
bones of many bird species persist into, but not
year, however (Flannery, island
through, archaeological zones showing human pres-
ence.
| will consider the Hawaiian. islands in detail.
We know 43 bird species only from their bones. Yet
bird bones are fragile and easily destroyed. We may
never find bones of all the now-extinct species, s
how many are missing? The bone record would be
complete only if all the recent species—those col-
lected or seen in the last two centuries—were also
found as bones. The proportion of recent species
also found as bones estimates how complete the
sample of species found only as bones is. The pro-
portion is about a half: across the Hawaiian islands.
we estimate there are about 40 species missing
from the record (Pimm et al., 1994). Add this num-
ber to the 43 known species extinctions and the
body count rises to 83.
James Cook found the Hawaiian islands in 1778.
International trade and colonization followed within
a generation. These new people cleared forests and
introduced cattle and goats. These destroyed native
plants as unprepared for large mammalian herbi-
vores as the birds were for the rats and. pigs the
Polynesians brought with them on their earlier col-
onization. Today, our only records of 18 species of
birds are the specimens collected by 19th century
naturalists. The body count rises to LOL.
What remains in the Hawaiian islands today?
Pimm et al. (1994) recorded that a dozen species
are so rare that there is little hope of saving them.
If we cannot find these species, then they probably
cannot find each other. A further dozen we can find,
but in numbers so small that their future survival
is uncertain. Of an estimated 136 species, only 11
survive in numbers that suggest a confident future.
Similar extinctions followed across the Pacific.
Over the roughly 1000 years, as the Polynesians
colonized the Pacific from New Zealand, north to
Hawai'i, and east to Easter Island, they extermi-
nated ~1000 species of birds or ~10% of the
world total (Pimm et al., 1994: Curnutt & Pimm.
2001). On some islands, they exterminated all the
bird species they encountered.
They extensively cleared lowland forests, espe-
cially the drier ones (and used only Stone-Age
technology to do so). Birds were not the only vic-
tims of these colonizations, incidentally. Of 980 na-
live Hawaiian plants, 84 are extinct and 133 have
wild populations of fewer than 100 individuals
Sohmer, 1994). These plant extinctions were the
consequence of recent human colonizations. Quite
how many plant species the Polynesians extermi-
nated we may never know.
Few species groups leave traces; land snails are
one and their losses illustrate the bizarre but ex-
tensive devastation that human actions can effect.
A predatory snail, Euglandina rosea, introduced to
many Pacific islands to control another introduced
snail, Achatina fulica, ate to extinction hundreds of
laxa of native Achatinella and Partula land snails
(Hadfield. 1986: WCMC., 1992).
(1 use the term
axa to include recognized geographically distinct
populations. Taxonomic uncertainties often raise
and sink their specific status. For those that are
now exlinct we may never resolve the issue.)
Nor are Pacific islands unusual in their species
losses. As European explorers moved from their
coastal waters from the early 1500s, Mauritius, Ro-
drigues, and Réunion in the Indian Ocean lost 33
species of birds, including the dodo, 30 species of
land snails, and 11 reptiles. St. Helena and Ma-
deira in the Atlantic Ocean have lost 36 species of
land snails (WEMC, 1992).
These examples raises two obvious questions.
The first is whether their numbers are unusual or,
192
Annals of the
Missouri Botanical Garden
alternatively, how many species should we expect
to go extinct each year? The background, that is,
pre-human rates of extinctions fluctuate consider-
ably over time and surely vary from one species
group to the next. However, a convenient (and like-
ly conservative) background rate of extinction is
about (May et al., 1995; Pimm
1995). Only about one in a million species
“one in a million”
et al.,
should terminate their existence naturally within a
year. The bird extinction rate is closer to one bird
species per year from a sample pool of only 10,000
bird species. This means that bird extinctions are
running 100 times the expected rate.
The second obvious question: Do we find evi-
dence of massive extinctions only on islands?
THE Dopo WENT ExriNcr (Too Bab!)
If island species were the only ones at risk, then
we consider their loss to be unfortunate, but relax
in the confidence that they were especially vulner-
able. This argument fails at two levels. The first
failure is that high rates of extinctions occur in
places other than islands. Here are three examples:
(1) A distinct and unusual flora defines the Cape
Floristic Region, which occupies a small area of the
southern tip of Africa. It comprises several vege-
tational types of which the fynbos is dominant in
area and contributes the most species. Of the Re-
gion’s 9030 species (Goldblatt & Manning, 2002
this issue), 36 species have become extinct in the
last century, and some 618 species are threatened
(Cowling, 1992). (I will always use "threatened"
a specific, technical sense to mean those species
thought likely to become extinct within at most a
few decades. Quite how long threatened species are
likely to last is a topic I discuss later.
(2) In. North America's rivers, Williams et al
(1992) described the mussels and clams in the Mis-
sissippi and St. Lawrence river basins. Of the 297
North American taxa of the two families Unionidae
and Margeritifidae, an estimated 21 have likely
gone extinct since the end of the last century. An-
other 120 taxa are threatened. Miller et al. (1989)
found that of ~950 taxa of freshwater fish in the
United States,
come extinct in the last 100 years. Northern lakes,
Canada, and Mexico, 40 have be-
southern streams, wetlands, and desert springs are
very different habitats, vet all have lost species.
(3) Of the 60 species of recent mammal extinc-
tions. worldwide, 19 are from Caribbean islands
(WCMC, 1992).
extinction rates of islands, and I will not consider
them further. Interestingly, 18 more were їп Aus-
tralia (WCMC, 1992), representing ~6% of its non-
This repeats the pattern of high
marine mammal species. The extinctions have been
equally divided between the southern arid zone
a sparsely inhabited area of mostly spinifex desert
and extensive pastoralism—and the wheat belt of
the southern tip of Western Australia—where 95%
of the natural woodland has been cleared (Short &
Smith, 1994). Another 43 Australian mammal spe-
cies have been lost from more than half of their
former ranges or survive on protected offshore is-
lands (Burbidge & McKenzie, 1989).
These examples refute the criticism that extinc-
tions are restricted to islands. In reviewing these
and other examples | am struck by the diversity of
taxa and ecosystems they encompass. Across these
examples, indeed for all well-known taxa, recent
extinction rates are 100 to 1000 times the expected
rate (Pimm et al., 1995; Lawton & May, 1995).
The second failure of the “it’s just islands” crit-
icism is more surprising. Certainly, greater numbers
and greater fractions of recent species extinctions
have been on islands than on continents. For in-
stance, since 1600, 97 of the total 108 bird extinc-
1994).
sland biotas are uniquely vulnerable to the human
tions have been on islands (Collar et al.,
introduction of previously absent herbivores, pred-
alors, diseases, and other natural enemies (Pimm,
1991). Species on continents are not so ecologically
naive.
There is another major factor that determines
threat. Most threatened species have small geo-
1998) and island
species’ ranges are inevitably smaller than conti-
graphic ranges (Stattersfield et al.,
nental ones. For a given range size, how do the
island and continental fractions of threatened spe-
cies compare?
Manne et al. (1999) calculated the ranges of all
the passerine birds in the Americas and their as-
sociated islands. (They comprise roughly a quarter
of all bird species.) To separate the effects of range
size, and island versus continental distribution, we
calculated the breeding range—henceforth, just
“range.” Some of the continental species inhabit
montane habitat “islands” isolated by a “sea” of
lowland habitats. We ask whether these montane
species suffer different levels of threat and so sep-
arate them from lowland species.
showed that for the 14 lowland, 8
montane, and 27 island species with ranges smaller
Manne et al.
than 1000 km? the proportions of threatened spe-
cies are high, but uncertain because of the small
sample sizes. Despite these uncertainties, for these
small ranges there is no evidence that island spe-
cies are more vulnerable than those on continents.
Perhaps one should expect this. Tiny ranges should
Volume 89, Number 2
2002
Pimm 193
Species Extinction
make species vulnerable to habitat loss, hunting,
and other threats wherever the species live.
Their most surprising conclusion emerges for
range sizes between 1000 km? and 100,000 km’.
Much smaller fractions of montane and island spe-
cies are threatened than of lowland species. At
ranges larger than 100,000 km? the proportions are
uniformly small in all three groups.
This unexpected result at intermediate ranges
has several possible explanations, but we consider
that local abundance is the most likely (Manne &
Pimm, 2001). We find that island species with a
range of (say) 10,000 km? are often locally abun-
dant on their island. Montane species with small
ranges are also locally numerous within their rang-
. These examples of numerous species with small
ranges have no match in continental lowlands.
There, species with such small ranges are almost
always very rare within those ranges (Brown, 1984;
Gaston et al., 1997). A reasonable explanation for
the abundance of island species is competitive re-
lease (MacArthur et al., 1972).
petitors, island species are able to attain. higher
With fewer com-
densities and are thus less likely to be threatened.
In sum, corrected for range size, continental spe-
cies are more—not less—likely to be threatened.
This unexpected vulnerability of continental. spe-
cies offsets their putative ecological sophistication
and experience of predators and other threats.
There seems to be no reason why continental spe-
cies will be spared the high rates of extinction hu-
manity first vested on insular species.
AS AMERICANS
HAPPEN
3. TRASH THE RAINFORESTS JUST
Dip THEIR FORESTS: NOTHING WILL
Extensive reductions in the forests of eastern
North
Surprisingly, only four bird species went extinct:
America occurred during the 19th century.
passenger pigeon, Carolina parakeet, ivory-billed
woodpecker, and Bachman’s warbler from reasons
wholly or mostly from habitat loss. Birds are well-
known, so we cannot plead ignorance of their ex-
tinctions. Critics use this apparent discrepancy to
claim that fears about massive global extinctions
(Budi-
1994). Just how many species should have
based on habitat losses are “simply wrong”
ansky,
gone extinct as a consequence of the loss of forests?
The answer comes from an extension of one of
the most well calibrated ecological relationships
1995).
quently provides a good description of the relation-
(Rosenzweig, The function, 5 = fre-
ship between the size of an area, A, and the number
of species, S, that it contains; ¢ and z are constants.
For real islands surrounded by sea, z is usually
about !4 (Rosenzweig, 1995).
“islands”
We can
then predict the reduction in numbers of species
Srinat OES
from A
Now, suppose we extend this to forest *
that remain amid a “sea” of deforestation.
„и аз the habitat's area is reduced
from
—the original extent of forest—to A
original пой
the area that now remains as forest “islands.” The
proportion of species lost = should ۴
(A JA usu). Thus S,
equals Sigina (Ano A
and the number of extinctions. S.
и ongin
euina equals Signa
— Su. Notice that we need an satiate of the —
of z, bil not of c. Does this recipe work or are one
or more of its assumptions flawed?
In North America, some 48% of the area covered
by the eastern forest at the time of European set-
tlement (1620) was still wooded at the point of its
lowest forest cover (roughly 1872: Pimm & skins.
1995). With A,4:/A,,, = 0.48 and z = 0.25, we
predict that — 179€ of the region's 160 forest birds
(27 species) should have become extinct. It is this
than the four
prediction, some six times greater
well-documented extinctions, that causes contro-
versy.
Does this discrepancy cast doubt on the predic-
tions of species losses from habitat reduction? It
does not. Those who point to the small number of
observed extinctions in the eastern forests mean
global extinctions—species that are lost every-
where. The prediction of 27 extinctions is based on
the number of species lost only within the region.
Some of these 160 bird species would survive even
if all the eastern forests were cleared. Their distri-
butions across the boreal forests of Canada or into
Central America would afford them a refuge while
U.S. forests were cleared.
How many species could become globally extinct
if all the eastern forests were felled? Which species
are found only in these forests, that is, how many
species are endemic to them? The answer is only
28. Now 17% of 2 : 4.76
roughly еи of a species higher than the
This prediction is
number of extinctions observed. | will not push my
luck
woodpecker, Picoides borealis, is three-quarters of
o argue that the endangered. red-cockaded
its way to extinction. The observed and predicted
numbers are remarkably close. This case history is
not the counterexample critics claim it to be. North
America lost few species because it had relatively
few species to lose
What happens in areas of the world that stand
lose many species? My colleagues and | have
applied this recipe to two such areas. The first is
1997, 1999a).
The region comprises four archipelagos: the Phil-
insular Southeast Asia (Brooks et al..
194
Annals of the
Missouri Botanical Garden
ippines, the Greater Sundas (Java, Sumatra, and
Borneo), northern Wallacea (Sulawesi and the Mo-
luccas), and the Lesser Sundas. Their forests hold
585 endemic species of bird—roughly 20 times
in half the area.
About 10% of the original area is cleared per de-
that of America’s eastern forest,
cade. Most of this deforestation has occurred re-
cently and ~60% of the original area is still for-
ested. Unlike the previous example, deforestation
has not yet caused any confirmed bird extinctions
in insular Southeast Asia. Extinctions take time fol-
lowing habitat loss, a point to which I must return.
hat does the species-area recipe predict about
the details of where extinctions will eventually oc-
cur? Across the region, some areas still have most
of their forests: Borneo had ~67%, for example,
when we assembled the forest cover data. (Forests
are shrinking rapidly, however.) Other areas have
almost none: Cebu, in the Philippines has < 1%.
And some areas have more endemic species than
others. Using the recipe, we predicted the numbers
of threatened bird and mammal species in each of
these four archipelagos, island by island. With a
few, interesting exceptions, there is a statistically
striking correspondence between the numbers of
species we predict should go extinct and those that
are currently threatened. Borneo, for example, has
38 endemics of which only 3 are considered to be
threatened: the recipe predicts 4. Sulawesi and as-
sociated islands also have about two-thirds of their
forest remaining, but there are 146 endemic spe-
cies strewn across these islands. The recipe pre-
dicts 14 should be threatened with extinction and
16 actually are. In contrast, in the Philippines, the
islands of Mindoro and the western and central Vi-
sayas have 19 endemic species; all are threatened,
while the recipe only predicts that 10 should be.
Where the recipe fails it usually does so by under-
estimating the number of threatened species: when
little forest remains, other factors—including hunt-
ing and invasive species—add to the threats.
The second area is the Atlantic coast forest of
Brazil (Brooks & Balmford, 1996). It has 214 en-
demic bird species. The area has four major sub-
divisions and for each there is a close match be-
tween the numbers of threatened species and those
we predict should become extinct solely on the ba-
sis of habitat loss. (The lowland forests are partic-
ularly hard hit, with only 296 of the forest remain-
11 endemic
species should be threatened; 9 are threatened.)
ing; the recipe predicts that 7 of
In sum, we have a well-calibrated ecological re-
lationship that predicts how many species should
become extinct following the loss of habitat across
three continents. Given enough time for the species
to die, as in North America, the predictions are
supported. Worldwide, for every extinct species of
bird there are 10 that are threatened. We predict
these much larger numbers, too, from the loss of
habitat in endemic-rich parts of South America and
tropical Asia. But we are still left with the criticism
that the species have not yet expired. There is a
lingering uncertainty that perhaps our worst fears
will not be realized. That leads to the final criti-
cism.
4. WHERE ARE THE BODIES?
If we are in the midst of an extinction crisis, why
are more species not going extinct? The reply is
that it takes time for (metaphorically) fatally
wounded species to expire. The point is made by
the extinctions of birds in eastern North America.
he low point of forest cover for these forests was
about 1870; the four fatally wounded birds lingered
for several decades, perhaps even a century, before
finally expiring.
This “many decades” matches many other sourc-
es of information. It fits with the IUCN definition
of “threatened”—a widely held expert opinion that
threatened species will likely go extinct within a
few decades. And it fits exactly with the few studies
that have explicitly examined forest fragments and
watched how fast species disappear from them
(Brooks et al., 1999b; Pimm & Brooks, 2000)
These studies suggest a species survivorship curve
with a half-life of roughly 50 years. That is, half
the species that will eventually expire do so within
the first 50 years, half of what remain expire in the
next 50 years, and so on. Given these results, over
what time period might the pending massive loss
of species from human actions unfold?
Pimm and Raven (2000) provided several an-
swers. The first comes from considering the large
fraction of species living within tropical forests and
how fast those forests are shrinking. A second an-
swer comes from looking at the hotspots—such
places as the Atlantic coast forests of Brazil and
Southeast Asia where endemic species are partic-
ularly concentrated.
About two-thirds of all species occur in the trop-
ics, most of them in tropical humid forests (Raven,
1980). Such forests include both evergreen rainfo-
rests and more seasonal ones. They originally cov-
ered from 14 to 18 million km?, depending on the
exact definition, and about half the original area
remains (Skole & Tucker, 1993). Much of the forest
reduction is recent, and clearing now eliminates
about 1 million km? of tropical forest in 5-10 years.
Burning and selective logging severely damages
Volume 89, Number 2
2002
Pimm
Species Extinction
several times the area that is wae چ (Nepstead et
al., 1999; Cochrane et al., 199€
To convert habitat loss to species loss, one ex-
tends the species-area relationship derived for is-
lands to predict how many species will not survive
” that remain amid a
in habitat fragment “islands
as described above. Then
“sea” of converted land
one updates the numbers each year as the total
forested area shrinks. Species that are classified as
threatened. will expire in decades to come and they
will be joined by other species for which we are
only now destroying their habitats. The doomed
species do not all die at once, but are spread over
time as determined by the species survivorship
curve. Combining these results gives an extinction
curve that I view as no more than a first sketch that
captures a few salient features.
Because the species-area curve is non-linear, the
clearing to date of half the humid forests should
have fatally wounded 15% of their species. This is
the case. Some 12% of all plants are threatened
(Walter & Gillett, 1998).
be an underestimate since many rare species hav e
This estimate is likely to
vet to be described. Of course, clearing the re-
maining half of these forests would eliminate the
other 85% of their species. Thus, the numbers of
fatally wounded species should accelerate. rapidly
to a peak by mid century. They will be joined by
ever-larger fractions of species jeopardized by the
interaction between the assumed constant rate of
forest clearing and the non-linear species-area
curve
The relative height of the peak depends critically
on the fraction of habitat that remains.
5% would protect about 50% of all the forests’ spe-
cies. Smaller percentages of remaining forest would
lead to very much smaller estimates of surviving
species. (About 5% of the world’s land surface is
protected at present, but that percentage includes
disproportionately large areas of desert and tundra
ecosystems. Protecting 5% of tropical forests will
require a considerable effort.)
that
The time delays before extinction mean
there will be far fewer species going extinct at pre-
sent than are being fatally wounded. The mode
predicts that current extinction rates should be
modest—on the order of a hundred species per
year, per million species. This matches current es-
timates (Pimm et al., There are as many
bodies as we expect, not far fewer. Extinction num-
bers will also peak in mid century, but will be
spread out over a century or more thereafter.
Modest tinkering with parameters does not alter
the “fewer extinctions now, many more later” fea-
ture of this curve, but the contribution of Myers el
A value of
does. They show that roughly 30 to 50%
al. (2000
of plant, amphibian, reptile, mammal, and bird spe-
—
cies occur in 25 hotspots that individually are no
more than 2% of the ice-free land surface. These
diverse taxa demonstrate that species with small
ranges are numerous and they are extraordinarily
concentrated. Nature has put her terrestrial species
in relatively few baskets. The sample applies to the
oceans: fishes and other organisms dependent on
coral reefs are similarly concentrated (Mc Allister et
al., 1994).
Myers el a
(2000) showed that human impacts
are malevolent, not random. Across the 25 hotspots,
an average of 12% of the original primary vegeta-
tion remains. This percentage should be compared
to the roughly 50% for tropical forests as a whole.
Even within the hotspots, Myers et al. found that
the areas richest in endemic plant species have
proportionately the least remaining vegetation and
the smallest areas currently protected (Fig. 1).
A second way to sketch the unfolding extinction
assumes that conservation actions immediately pro-
tect all the remaining habitat areas within the hot-
spots. Applying the species-area curve to the in-
dividual hotspots predicts that 18% of all their
species would eventually go extinct. [Since Myers
et al. (2000) showed that hotspots hold 30-50% of
all species, see above, this percentage is also con-
sistent with the fraction of currently threatened spe-
cies.| Yet another sketch assumes that the hotspots
higher than global average rate of habitat loss con-
tinues for another decade until only the areas cur-
rently protected remain. The hotspots would even-
tually lose 44% of all their species (Pimm & Raven.
2000).
None of these three sketches captures the inad-
equacy of some of the protected areas. the so-called
"paper parks." Nor do these ideas consider the
added threat of global warming that will doubtless
limit the effectiveness of sharply delimited, small
reserves. Also excluded are the major threats that
and weedy species
invasive s
pose to the remaining species. Often listed as the
most important factor in causing threat and extinc-
tion. the impacts of invasive species on islands are
Plant
introductions are a major threat to the Cape Region
of South Africa, 1992).
The distinction between these three sketches is
well known. Continents. are. vulnerable. too.
for instance (Cowling.
artificial. Many species live in tropical forests that
are also hotspots. Yet others live in tropical forests
that are not and some live in hotspots that are not
tropical forests. Nonetheless. the sketches capture
views of the size and time-scale of the processes
driving the unfolding extinction.
196
Annals of the
Missouri Botanical Garden
50
yo]
2
$ 404
О
o
>
a
Е 304 ш
о © ga
3 1 = m
Ф E ш
НЕ 201 > шш.
a B
£ ©
s 74 m^" E g :
E и
S БЕ a = > ©
as © HB о
59 o m o o
bd T Y LI 1
0 5000 10000 15000 20000 25000
Number of endemic plant species
Figure 1.
The percentage of the original habitat remaining and the percentage of the original habitat protected are
сона smaller іп those hotspots that contain the greatest number of endemic plant species. From data in Myers et
al. (200
The first process is the rapidly accelerating loss
of presently extensive, but rapidly shrinking, trop-
ical forests. Protecting substantial and representa-
tive areas requires prompt action. This is unlikely
happen unless industrialized nations become
more deeply involved with funding conservation in
developing ones. Without such action, the loss of
species from these areas will overtake the loss of
species from hotspots within a few decades.
The second process is the rate of loss of species
from hotspots. Losses here should dominate for the
next few decades, since hotspots are already se-
verely fragmented. [By definition: Myers et al.
(2000) defined hotspots to have unusual numbers
of endemic species and to have suffered dispropor-
tionate habitat losses.] Only immediate conserva-
tion actions, including restoration of damaged hab-
itats, can prevent further species loss. And unless
there is immediate action to salvage the remaining
unprotected areas, the species losses will more than
double. As Myers et al. pointed out, the current
unprotected areas constitute only a little more than
l million km?. High concentrations of small-ranged
species make many species vulnerable, but equally
they permit a concerted effort to prevent future ex-
tinctions.
CONCLUSIONS
The dodo did not go extinct. Humanity blud-
geoned it into oblivion. With it went 10% of the
planet’s birds and, in all probability, similar frac-
tions of other poorly known species of plants and
animals. That we did not identify and name all the
species that disappeared is not a credible argument
or their continued survival. The Vietnam memorial
on the Mall in Washington, D.C., is a poignant list
of all the Americans who died in the U.S.A.’s war
in that country. A far smaller list of names appears
on a memorial in the village in England where I
was born to men who died in France between 1914
to 1918. I recognize those names as just a sample
and, relative to the village’s small population, read-
ily extrapolate to the massive slaughter of men
across the entire country. While a complete list of
extinct species would be useful, it is not essential
to perceive or to estimate the size of the current
crisis
най the threats of future extinctions from
the few that have occurred in North America is
likewise the consequence of misinterpretation. Most
of the recent known bird extinctions on continents
happened in North America following European
Volume 89, Number 2
2002
Pimm 197
Species Extinction
colonization. Quite what happened in Europe when
its forests were cleared centuries earlier we may
never know. Consequently, North America is the
crucial case history of forest and species loss. It
teaches that we lost 4 of 28 of its endemic forest
bird species, almost exactly what the species-to-
area calibrations predict on the basis of a 50% re-
duction of forests. (Three more species were hunted
to extinction: the great auk, the Labrador duck, and
the Eskimo curlew.) The major tropical forests in
the Amazon, the Congo, and New Guinea have al-
ready lost half their area, are shrinking by the day.
and yet they hold more than 10 times the number
of bird species that were found in eastern North
America. The hotspots are already depleted even
further. The North American case history is most
telling when scaled appropriately.
Some scientists have overestimated the numbers
of species that should be going extinct per year at
present. The fault lies solely with the assumption
that species would die out immediately. Some do,
but most manage to linger. We have yet to realize
the 10% loss of species—roughly the fraction of
because
well-known species that are threatened
the destruction of the most species-rich ecosystems
has only unfolded in the last half century. Yet this
overestimation is simply fixed by changing the text
from predictions of “actual extinctions” to predic-
tions of species “being on an inexorable path to
extinction.” Unless we protect more of the planet’s
remaining natural areas, by the end of this century
that distinction will seem absurdly trivial.
Literature Cited
Barney, G. О. x The Global 2000 Report to the Pres-
ident, Vol. 2 . Government Printing Office; Wash-
— D.C.
Brooks, T. A. cape
tinctions. Nature 380:
S, Pimm & N. E: Collar. 1997.
f threatene n
Conse rvation Biol. 382-38
— — apos & C. FK: — 1999a. Threat
from dota to montane and lowland birds and
mammals in insular Southeast Asia. J. Animal Ecol. 6
1996. Atlantic Forest ex-
Deforestation
edie ts үз insular
— in
Southéast Asia.
1061-1078
———— & J. О. Ovugi. 1999b. Time lag between
deforestation and bird Pa * in tropical forest frag-
its. Conse — Biol. 13: 1140-115
Brown, J. Н. 1984. On the re — Бен
dane 'e and — of species
veen abun-
‚ Amer. Naturalist 124:
55
апей, 8, 1993. The d myths. U.S. News and
World Report, December 13, 1993, 81—83
—— ———. 1994. Extinction or miscalculation? Nature 370:
105,
Burbidge, „ McKenzie. 1989. Patterns in the
modern ox ci. of — Australia’s vertebrate fauna:
C
‘auses and conservation implications. Biol. Conserva-
tion 50: 143 pu
Сос ev. MA A. Alencar, М. D. Schulze, С. М. Souza.
Jr., a nem P. Lefebvre & E. A. ee 1999,
a feo ks in the fire — of closed canopy
— al forests. Science 284: 1832-183:
Collar. N., M. Crosby A. Statte si ld.
Watch 2. — Institution Press.
D.C.
Cowling, R. M. (editor). 1992. The
0: iv. Press, Cape Town.
Curnutt, J. & S. L. Pimm. 2001.
in Hawaiʻi and the E Pacific before first contact?
Pp. 15-30 in J. M. Scott, S. Conant & C. van Hiper IH
(editors), Eula P Conservation, and Man-
agement of Hawaiian non
Studies in — Ese
Н Байа extinction.
199 1. Birds to
Washington.
—
Ecology of Fynbos.
How many bird species
Vanishing Avifauna.
Flannery, i E Science 283:
182-
Gaston, a T Blackburn & J.
abundance-range
Lawton. 1997, Interspecific
size relations —— An appraisal of
mechanisms. J. eis Ecol. 66: 579-601.
Goldblatt, P. & J Manning. 2002. Plant diversity о
the Cz T = е southern Africa. Ann. Missouri Bot.
813 02.
. 1986. Extinction in Hawaiian achatinel-
Malas ologia 27: ool.
Lawton, J. H. & . Extinction Rates. Ox-
ford Univ. Press. Oxfor
Lomborg. B. 2001. The Skeptical Environmentalist: Mea-
suring the Real State of the World. Cambridge Univ.
Press, Mey 7
MacArthur, I . J. M. Diamond & J. Karr.
compe nsation in island faur
» С. & M. L. Plummer.
fred A. Knopf, New York.
& S. L. Pimm. 2001.
rarity: Which species are threatene d and which will
next? — C — 4: 221-230.
1999, Relative risk
of extinction of passe rine p on continents and is
;. Nature 399: 258-201.
May. R. M., J. H. Lawton & N. E. Stork. ‚ Assessing
extine tion rates. Pp. 1-24 in J. H. — R. M. May
С Extinc а Rates. КИ Univ.
Schueler, C.
A apping and GIS — of the global
dais hr nsity
nas. Ecology 53: 342
1995. Noah’s Eget м.
Beyond eight forms of
. Brooks € S. L. Pimm.
distribution of coral reef fishes on an equal-area grid.
Pp. 155-17 Miller (editor). Advances in Map-
ping the — of Nature. Chapman & Hall, London.
Miller, R. R., J. D. Williams & E. Williams. 1989,
Extinctions of North American fishes during the past
Fisheries 14: 22-38.
yers, N., iae p C. G. Mittermeier, G. A. B.
da onsena & J. Kent. 2000. Biodive 2: — for
conservation Priorities. Nature 403: 853-
epstead, i. A. Verssimo, A. Aler
Lima, P. Lafe hië, P. Schlesinger, C. Pete P. Moutin-
Mendoza, M. Cochrane € V. Brooks. 1999,
Large-scale a Ren us as forests by
logging and z ^s. Nature 398: )
The lance of — ? Univ. Chicago
century.
di TM E.
)5—508.
. The World ч to Pimm: A Scientist
Audits he Кан, McGraw Hill, York
Askins. 1995. Forest um s predict bird ex-
198
Annals of the
Missouri Botanical Garden
tinctions in eastern North America. Proc. Natl. Acad.
Sci. U.S.A. 92: 9343-9347.
& Т. М. Brooks. 2000. The — Extine tion: How
large, how soon, and where? Pp. in P. Raven
(editor), Nature and Human Society: me Quest for a
Sustainable World. National Academy Press, Washing-
ton,
& P H. €— 2000. Extinction by numbers. Na-
ture 403: 843—
———, М. P. viral & L. J Justice. 1994. Bird extinc-
tions in the pop" Pacific. Philos. Trans. Roy. Soc
London B344: 27—
— ———, J. Russell, n (а Gittleman & T. М Brooks. 1995.
The future of biodiversity. Science 269: 347-350.
Raven, P. H. (editor). 1980. Research Priorities in Tropical
Biology. National Academy of Sciences Press, Washing-
ton,
‚ 1988. Our араа — al forests. Pp. 119—
122 in К. О. Wilsor 'r (editors), Biodiver-
sity, National Ac a my Press, Washington, D.C,
Rosenzweig, М. L. 1995. Species Diversity in Space and
Time. Cambridge Univ. Press, Cambridge, U
Short, J. & A. Smith. 1994. Mammal — and recovery
in car J. Mammalogy 7:
Simon, J. & А. Wildavsky 1993.
are — = York Times 142 May 13, : A23.
Skole, D. & C. J. Tucker. 1993. Tropical — and
one b — ntation in the — Satellite data from
1978 to 1988. Science 260: 1905-1910.
Sohmer, n 1994. Pp. 43-51 de. .-1. &
rial Eco
(editors), Biodiversity and Terre stems,
Monograph Series Vol. 14. E of Шы e ade-
mia Sinica, Taipei.
piy A. . Crosby, A. J. Long & D. C. We
8. Endemic Bird Areas of the World: Priority ы,
or — Conservation. BirdLife Conservation Se-
7. Cam ric
Sie aan: D. . 1995. ride extinctions of Pacific
— d „Biodiversity meets zooarchaeology. Sci-
e 261:
Si. P i ‘Dopieal Rain Forest: A Political Ecology
ji еш пш Studies in the Environment
15, Institute of Economic Affairs, London.
Wa * K. S. & н. J. Gillett. 1998. IUCN Red List of
Threatened deis E. oe World Conservation
Jnion, Gland, Switz
WCMC. 1992. Global а Status of the Earth’s
Living Resources. World Conservation Monitoring Cen-
tre, Chapman & Hall, Londor
Williams, J. D., M Warren, J ‘ummings, J. L.
Harris & R. J. Neves. 19€ servation status of
freshwater mussels of the da oy ач and Canada.
isheries 18: 6-22
Wilson, E. О. 1988. Pp. 3-18 in E. O. Wilson € E
Pet ا F: тт National Academy Press,
Washington. D.
THE GLOBAL 200: PRIORITY David M. Olson?’ and Eric Dinerstein?
ECOREGIONS FOR GLOBAL
CONSERVATION!
ABSTRACT
A global strategy to conserve biodiversity must aim to protect represe ntative examples of all of the world’s ecosystems,
as e as those areas that zc exc eptional concentrations of species and endemics. Although lacking the richness
of tropical forests, deserts, tropical lakes, and subpolar seas all contain distinct species, communities, and ecological
phenomena. We analyzed global — of biodiversity to identify a set of the Earth’s terrestrial, freshwater, and marine
ecoregions that harbor exceptional biodiversity and are representative of its ecosystems. We placed each of the Earth’s
ecoregions within a system of 30 biomes and biogeographic realms to facilitate a re — — Biodiversity
features were compare E — ecoregions to assess their irreplaceability or pere tive These features included
species richness, endemic species, пеем higher taxa, unusual ecologic al or a en phe nomena, * = global
rarity of habitats. This | process yielded 238 ecoregions—the Global 200—compis] of 142 terrestrial, 53 freshwater,
and 43 marine priority ecoregions. Effective conservation in this set of ecoregions uña help conserve ia most
outstanding and representative habitats for biodiversity on this planet.
Key words: biodiversity, conservation, ecoregions, endemism. global, phenomena, priority-setting, representation.
Tropical rain forests rightfully receive much con-
~
yr those with unusual ecological or evolutionary
servation attention as they may contain half of the phenomena. We must also target representative ex-
world’s species. A comprehensive strategy for con- amples of all of the world’s biomes within each bio-
serving global biodiversity, however, must strive to geographic realm where they occur (Fig. 1). Be-
save the other 50 percent of species and the dis- cause of distinct biogeographic histories, similar
tinctive ecosystems that support them. For example. kinds of ecosystems found on different continents
while they may not support the rich communities or in different ocean basins support unique assem-
seen in tropical rain forests or coral reefs, tropical — blages of species and higher taxa. For this reason,
dry forests, tundra, polar seas, and mangroves all global strategies should strive to conserve examples
harbor unique species. communities, adaptations, of every biome in each realm where it occurs for
and phenomena. Some of these biomes, such as terrestrial, freshwater, and marine biodiversity (Ol-
tropical dry forests and Mediterranean-climate son & Dinerstein, 1998; Udvardy, 1975: Dasmann.
shrublands, are more threatened than are tropical 1974). Here we present the Global 200—an al-
rain forests and require immediate conservation ac- tempt to identify a set of ecoregions whose conser-
tion. To lose examples of these assemblages would vation would achieve this goal of saving a broad
represent an enormous loss of global biodiversity. diversity of the Earth's ecosystems (Figs. 2. 3). This
Limited funding compels the conservation com- paper expands and updates an earlier analysis by
munity to be strategic and earmark the greatest Olson. and Dinerstein (1998). Several additional
amount of resources to protect the most outstanding ecoregions have been identified through ongoing re-
and representative areas for biodiversity. On a glob- gional analyses (e.g.. Wikramanayake et al.. 2001)
al scale. this requires identifying large regions with and the marine Global 200 have been reduced.
exceptional levels of species richness or endemism, largely due to combining several adjacent areas
! We thank the regional experts, biologists, — conservationists who contributed their time and knowledge to the
conservation analyses that went into the Global 200. J. Leape and С. Hails have provided critical — for this
effort. The staff of WWE contributed greatly to A — assessments from which the map is derived. We thank
World Wildlife Fund’s Conservation - ience P rogram for their contribution 2 the analysis and preparation of is Global
200, specifically R. Abell, T. Allnutt, C. Carpe F p — o, P. Hu rley ‚ Kassem, Н. Strand, M. Taye, M. Thieme,
W. Wettengel, E. Unde — E. W TS . Itoua, C. Loucks,
Y. Kura, J. Morrison, and G. Powell. J. Martin- Po anc Ü . Lagler he [pedi in many ways to facilitate the completion
of this project. We thank the staff of pad WWF Network, including all of the ont! organizations, various field offices
and programs, and associates, for their review and comments on си "dra
? Conservation Science Program, World Wildlife Fund-US, 1250 24th — NW, Washington,
dolson@wes.org, eric.dinerstein@ww
3 WCS South Pacific Program, У Box. 3080. Lami, Fiji.
=
=
Ф
7
Es
=
تع
=
3
T
DI
x
E
-
=!
—
S
a
).C. 20037, U.S.A.
ANN. Missouni Bor. GARD. 89: 199-224. 2002.
200
Annals of the
Missouri Botanical Garden
Oceania
—
p Tropical — зазора Moist Broadleaf Forests
cal Dry B
4
Pis
ك
d
qu Шере vi nega Grasslands, Savannas,
and Shr
Figure 1. Terre
(sensu Dasmann, 1974; Udvardy,
—
into larger units. In addition, the conservation sta-
tus of all ecoregions has been estimatec
Ecoregions are regional-scale (continental-scale)
units of biodiversity. We define ecoregions as a rel-
atively large area of land or water containing a
characteristic set of natural communities that share
a large majority of their species, ecological dynam-
ics, and environmental conditions (Dinerstein et al.,
1995; Groves et al., 2000). They function effective-
ly as coarse-scale conservation units because they
and
their extent roughly coincides with the area over
which key ecological processes interact most
strongly (Orians, 1993).
For each of the Earth's 30 terrestrial, freshwater,
encompass similar biological communities,
and marine biomes (formerly referred to as major
habitat types in our previous analysis), we com-
pared the biodiversity of each constituent. ecore-
gion. Those ecoregions whose levels of biodiversity
were considered exceptional (that is, highly dis-
tinctive or irreplaceable;
1995; Pressey et al.,
see Dinerstein et al.,
1994) for their biome, or
which were considered the best example of a biome
within a realm (even if none of the candidates har-
bored exceptional biodiversity), were identified as
areas of particular importance for achieving global
conservation goals, This prioritization yielded 238
ecoregions—the Global 200—comprised of 142
С Temperate Grasslands, Savannas, and Shrublands
ШЕШ Flooded Grasslands and Savannas
WÎ Montane Grasslands and Shrublands
И Tundra
IN Mediterranean Forests, Woodlands, and Scrub
ШЕШ Deserts = Xeric Shrublands
ШШ Mangrov
sstrial biomes, terrestrial and freshwater biogeographic realms, and marine biogeographic realms
975
terrestrial, 53 freshwater, and 43 marine ecoregions
nested within 30 biomes and 8 terrestrial and fresh-
water biogeographic realms and 13 marine biogeo-
graphic subdivisions (Table 1)
DELINEATION OF ECOREGIONS
TERRESTRIAL ECOREGIONS
Dasmann (1974) and Udvardy (1975) were the
first to conduct a global representation analysis for
terrestrial conservation. Dasmann's system of 198
biotic provinces and Udvardy's 193 units are nested
within 7 biogeographic realms and 13 terrestrial
biomes and 1 freshwater biome. Both these geo-
graphic models serve as logical frameworks for
analyses of global representation.
and Udvar-
dy's biotic provinces, however, limits their utility as
The relative coarseness of Dasmann's
regional conservation planning tools as many dis-
tinctive biotas may remain unrecognized. We be-
lieved a more finely resolved map of biodiversity
patterns was required, one that mapped distinctive
biotas within single, continuous biomes. This called
for intensive regional analyses of biodiversity pat-
terns across five continents by synthesizing existing
201
Olson & Dinerstein
The Global 200
Volume 89, Number 2
2002
әче, ur paist]
SUOIBAIOI9 ou 0] puodsoiio) ләш up "Jn[e^ X022 QOS 1694019) [8459.19] əy С 21n31 y
aAejuoasoadoa pue soinjeoj Nprsaoarporq aurpue]sjno цим SUDIDAIOIA ӘЗІР] suor.
spuejqniys pue
se^ojbuew EN 'seuue^es "spue|sseJc једіодапѕ pue үедіо К
spue|qnaus эпәх pue sueseg Ш
qnaos pue 'spue|pooM 's153104 ueaue1123!PaN ни
spuejqnius pue spue|sseJc әйезиору аш
SeuueAePs рие spue|sseJ9 рәрооу
spuejqniys pue 'seuueAes 'spue|sseJ) ajejeduijg| — —
202
Annals of the
Missouri Botanical Garden
Table 1.
and biogeographic realms
Global 200 ecoregions organized by biomes
. the estimated conservation sta-
low ritical
and RS dor relatively
arked by asterisks (*) rep-
tus for each ecoregion is noted as fol ¿ for сг
or endangered, V for vulnerable,
e or intact. Ecoregions m
areas presently under review for elevation to
С Siobal 200 status based on their biodiversity features and
representation value.
TERRESTRIAL REALM
TROPICAL AND SUBTROPICAL MOIST BROADLEAF FORESTS
Afrotropical
Guinean Moist Fore CE
2. Congolian Coastal For 's CE
3. Cameroon Highland — CE
4. Northeastern Congo Basin Moist em V
5. — Congo Basin Moist Fores RS
6. Western Congo Basin Moist Sat V
Albertine Rift Montane Forests CE
‚ East African Coastal Forests CE
A Eastern Arc Montane Forests CE
10. Madagascar Forests and Shrublands CE
11. Seychelle and Mascarene Moist Forests CE
Australasian
12. Sulawesi Moist — CE
13. Moluccas Moist For ү
14. Southern New ме — Forests V
New Guinea Montane Forests RS
v Solomons- Vanuatu- н н Moist Forests V
. Queensland Tropical Fore V
tà New Caledonia Moist Ё CE
19. Lord а Norfolk энге Torati CE
ме —
). lia Ghats — Forests CE
* Sri Lankan Moist Forests CE
22. сое Indochina — ‘al Moist For-
ү
23. — China-Hainan Moist Forests CE
24. Taiwan Montane Forests ү
25. Annamite Range Moist Fores
26. Sumatran Islands Lowland and ‘Montane For-
ests CE
27. Philippine Moist m CE
28. Palawan Moist F CE
S Kayah- Karen/Tenasserim Moist Forest [
‚ Peninsular Malaysian Lowland and —
Forests V
3l. Borneo Lowland and Montane Forests CE
32. Nansei Shoto Archipelago Forests CE
" Eastern Deccan Plateau Moist Fore CE
34. Naga-Manupuri-Chin Hills — forests ү
35. = Mountains Moist Forest RS
36. Western Java Montane Forests CE
Neotropical
37. Greater Antillean Moist Forests CE
* (Lesser Antillean Moist Forests) CE
Talamanc an⸗ d Pacific Forests RS
E Chocó-Darién Moist Forests RS
40. Мору panne Montane Forests CE
41. Coastal Venezuela Montane Forests ү
42. uel Moist Forests RS
43. Napo Moist Fores V
44. Rio Negro-Juruá Moist Forests CE
45. Guayanan Highland Moist Forests RS
Table 1. Continued.
46. Central Andean Yun CE
47. Southwestern — Moist Forests RS
48. Atlantic Forests CE
Oceania
49. South Pacific Island Forests CE
50. Hawaii Moist Forests
TROPICAL AND SUBTROPICAL DRY BROADLEAF FORESTS
Afrotropica
51. Madagascar Dry Forests CE
ок
. Nusa Tenggara Dry Forests CE
. New Caledonia Dry Forests CE
i^ —
. Indoc hina Dry Forests CE
55. Chhota-Nagpur Dry Forests CE
Neot ropical
56. Mexican dry Fore CE
di Tumbesian- Andean ‘Valleys Dry Forests CE
C hiquitano Dry Forests CE
M Atlantic Dry Forests CE
Oceania
60. Hawaii Dry Forests CE
TROPICAL AND SUBTROPICAL CONIFEROUS FORESTS
Nearctic
61. Sierra Madre Oriental and Occidental Pine-
Oak Forests
Neotropical
62. Greater Antillean Pine Forests CE
63. Mesoamerican Pine-Oak Forests CE
TEMPERATE BROADLEAF AND MIXED FORESTS
Australasia
64. Eastern Australia Temperate Forests CE
65. Tasmanian Temperate Rain Forests ү
66. New nd Temperate Forests ү
m d
р Himalayan Broadleaf and Conifer
68. West stern n Himalayan Temperate Forests CE
Nearctic
69. Appalac hian and Mixed Mesophytic Forests V
Palearcti
di South China Temperate Forests V
. Russian Far East Temperate Forests V
TEMPERATE CONIFEROUS FORESTS
Nearctic
72. Pacific Temperate Rainforests CE
13. Klamath-Siskiyou Coniferous Forests CE
74. Sierra Nevada Coniferous Forests CE
ТӘ; — Coniferous and Broadleaf For-
(
КО. al
. Valdivian Temperate Rainforests/Juan Fer-
nández Islands CE
E — Islands and Desventuradas
slands CE
Paleare ис
77. European-Mediterranean Montane Mixed
Forests
Volume 89, Number 2
2002
Olson & Dinerstein 203
The Global 200
Table 1. Continued.
Table 1. Continued.
78. Caucasus-Anatolian-Hyrcanian Temperate
Orests
79. Altai-Sayan Montane Forest V
80. He ngduan Shan Coniferous Todd RS
BOREAL FORESTS/TAIGA
Nearctic
81. Muskwa/Slave Lake Boreal Forests RS
82. — Boreal Forests S
PEE
Ural Mountains Taiga V
^ Eastern Siberian Taiga RS
85. Kamchatka Taiga * Grasslands RS
TROPICAL AND SUBTROPICAL GRASSLANDS, SAVANNAS, AND
UBLANDS
Horn of — Acacia Savannas ү
87. East African Acacia Savannas \
88. тте al d Eastern Miombo Woodlands V
89. Sudanian Savannas CE
Australasia
90. Northern Australia and Trans-Fly Savannas RS
a“ Malayan
Ter
rai-Duar Savannas and Grasslands CE
ub us a
92. Llanos Savannas M
93. Cerrado Woodlands and Savannas \
TEMPERATE GRASSLANDS, SAVANNAS, AND SHRUBLANDS
Nearctic
94. Northern Prairie CE
* Ta * prairies
ls pow
95. peo Steppe CE
PED tic
. Daurian Steppe \
FLOODED GRASSLANDS AND SAVANNAS
Afrotropica
DE. S — Sahelian Flooded Grasslands and Sa
98. Zambe ian Flooded Savannas \
Indo-Malayan
‹ ann of Kutch Flooded Grasslands CE
Veotropica
100. Everglades Flooded Grassland \
101. Pantanal Flooded Savannas CE
MONTANE GRASSLANDS AND SHRUBLANDS
ji e al
2. Ethiopian Highlands CE
14 Southern Rift Montane Woodlands CF
104. East African Moorlands RS
105. Drakensberg Montane Shrublands and Wood-
СЕ
Australas
106. Central Range Subalpine Grasslands RS
ne Malay
Kinabalu Montane Shrublands RS
a ee
108. Rehet Andean Paramo RS
109. Central Andean Dry Puna ۷
Paleartic
110. Tibetan Plateau Sleppe ү
111. pires Asian Montane Steppe and Wood-
lar
112. * rn Himalayan Alpine Meadows RS
TUNDRA
Vearcti
113. AL askan North Slope Coastal Tundra RS
114. Canadian Low Arctic Tundra RS
Palearctic
115. Fenno-Scandia Alpine Tundra and Taiga V
116. Taimyr and Siberian Coastal Tundra RS
117. Chukote Coastal Tundra RS
MEDITERRANEAN FORESTS, WOODLANDS, AND SCRUB
l
Afr реа а
118. Fynbos СЕ
—
—119 ee stern Australia Forests and Scrub CE
120. Southern Australia Mallee and Woodlands C
Nearctic
121. California Chaparral and Woodlands CE
t
Veotropical
122. Chilean Matorral CE
Palearctic
123. Mediterranean | Forests,
Scrub
Woodlands. and
DESERTS AND XERIC SHRUBLANDS
Afrotropical
124. Namib-Karoo-Kaokoveld Deserts
125. Madagascar Spiny p” СЕ
126. Socotra Island Deser
127. Arabian ШМ Woodlands and Shrub-
ls \
lands
lustralasia
128. Carnavon Xeric Serub CE
129. Great Sandy-Tanami Deserts RS
Vearcti
130. Son ioran-Baja Des RS
131. Chihuahuan-Te өр icán Deserts \
Veotropical
132. Galapagos Islands Serub \
133. Atacama-Sechura Deserts V
Palearctic
134. Central Asian Deserts CE
MANGROVES
\frotropical Atlanti
135. Gulf of Gaines Mangroves CE
ыст E пар
African Mangroves СЕ
: 37. en ar Mangroves CE
Australasia
138. New Guinea Mangroves RS
Indo- Malayan Indo-Pacific
139. Sundarbans Mangroves CE
140. Greater Sundas Mangroves CE
Neotropical Atlantic
141. Guianan—Amazon Mangroves RS
e ta Pacific
142. Panama Bight Mangroves RS
204
Annals of the
Missouri Botanical Garden
Continued.
Table 1.
Continued.
Table 1.
FRESHWATER REALM
LARGE RIVERS
Afrotropical
143. Congo River and Flooded Forests
Indo-Malayan
144 Lm ‘kong River
Nea
45. pr Hiver
146. Lower Mississippi River
Neotropical
147. Amazon River and Flooded Forests
148. — o River and Flooded Forests
Palearct
149. Yangtze River and Lakes
LARGE RIVER HEADWATERS
— a
150. Congo Basin Piedmont Rivers and Streams
ата ic
Mississippi Piedmont Rivers and Streams
Veni al
2. Upper Amazon Rivers and —
СЯ Upper Paraná Rivers and Streams
154. Brazilian Shield {тэн иде Rivers and
Streams
ARGE RIVER DELTAS
ye
155. Mas — Delta
e Malay
Indus River Delta
— ic
157. Volga River Delta
158. Mesopotamian Delta and Marshes
159. Danube River Delta
160. Lena River Delta
SMALL RIVERS
— 'al
61. Upper Guinea icd and Streams
y Madagascar Freshw
163. Gulf of Guinea Rivers and Streams
64. Cape Rivers and Streams
nd
Ce
. New Guinea Rivers and Streams
Ti New Caledonia Rivers and Streams
167. Kimberley Rivers and Stream:
168. Southwest Australia Rivers and Streams
169. Eastern Australia Rivers and Streams
New Zealand Rivers and Streams)
ens Malayan
. Xi Jiang Rivers and Streams
| Western Ghats Rivers and Strea
172. Southwestern Sri Lanka Rivers and Streams
173. Salween River
174. Sundaland Rivers and Swamps
Nearctic
175. Southeastern Rivers and Streams
176. Pacific Coastal Rivers and Streams
177. Gulf of Alaska Coastal Rivers and Streams
Neotropical
178. Guianan Freshwater
179. Greater Antillean Freshwater
* (Southern Cone Freshwater, especially Val-
divian region)
* (Atlantic Coast rivers of SE Brazil, Uruguay)
Palearctic
180. Balkan Rivers and Streams
*
—
expansion to Mediterranean region in gen-
eral including western North Africa
181. Russian Far East Rivers and Wetlands
* (Aral Sea Basin, particularly Syr- and Amu-
Dar'ya Rivers)
ARGE LAK
пто uu
82. Rift Valley Lakes
Netopia l
83. High Andean Lakes
Pia
Lake Baikal
85. Lake Biwa
Afroropi al
86. — Crater Lakes
—
187. Lakes Kutubu and Sentani
188. Central Sulawesi Lakes
— Malayan
9. Philippines Freshwater
Ts Lake Inle
191. Yunnan Lakes and Streams
rei 'al
. Mexican Highland Lakes
XERIC BASINS
Australasia
adm — Australian Freshwater
Nearc
veg С hihuahuan Freshwater
— tic
95. Anatolian Freshwater
MARINE REALM
POLAR
Antarc
196. — ‘tic Peninsula and Weddell Sea
Arctic
197. Bering Sea
198. Barents-Kara Seas
TEMPERATE SHELF
ee
19¢ editerranean *
Poda Temperate Atlan
200. Northeast Мя Shelf Marine
201. — nks
202. Ches i
p. Temperate — Pac iſic
203. Yellow
AA meh Sea
AND SEAS
Volume 89, Number 2
2002
Olson & Dinerstein 205
The Global 200
Table 1. Continued.
Southern Ocean
205
200.
Patagonian Southwest Atlantic
Saam Australian Marine
New Zealand Marine
TEMPERATE UPWELLING
North
208
2]
= C
Te mperate Indo- Pacific
Agulhas Current
TROPICAL UPWELLING
Central indo- Pacific
=
stern — Marine
Eastern lids: Pac
213.
214
Panama B i
. Gulf of California
э. Galápagos Marine
Eastern Tropical Atlantic
216.
Canary Current
TROPICAL COR!
pn — Pac ific
222.
223.
el :
Nansei Shoto
New Caledonia Barrier Reef
Great Barrier Reef
Lord Howe-Norfolk Islands Marine
Palau Marine
Andaman Sea
fal astern Indo-Pacific
220.
228.
228.
=
233.
234.
w
=
Societies/Marquesas/Tuamotus Marine
Hawatian —
Nui Marit
Fiji Barrier Reel 8 Marine
We ste rn Indo-Pacific
Maldives. Chagos. Lakshadweep Atolls
East Afric ‘ап Marine
West Madagascar Marine
* (The Mascarene Islands are under consid-
eration due to high numbers of endemic reel
ish)
* (The Maldives are under consideration for
extension to include Sri Lanka and southern
Indian coast)
We stern — Atlantic
23:
236.
7. Southern Caribbean Sea
oamerican Ree
ost Antillean Marine
Northeast Brazil Shelf Marine
RS
ү
ENE
classifications from finer scales. Furthermore, de-
lineations were conducted in collaboration with
hundreds of regional experts and included exten-
sive literature reviews. The result is a digital map
of 867 terrestrial ecoregions, classified within bi-
omes and realms, to be used for prioritv-setting
analyses (Olson et al., 2001). This map provides a
much more detailed picture of how species assem-
blages are distributed across the world. The in-
creased resolution is most apparent in the tropics
where Dasmann (1974) and Udvardy (1975) iden-
tified 115 and 117 provinces respectively, com-
pared to 463 terrestrial ecoregions.
Dasmann and Udvardy both went on to assess
how well existing protected areas represented the
Earth's terrestrial biomes and realms. Biotic prov-
inces with little or no protection were identified as
priorities. The Global 200 analysis frames the goal
of prioritization differently: We ask which regions
should be a priority for conservation action (e.g..
designating and strengthening protected areas) be-
cause of their outstanding biodiversity features or
their representation value. We also apply this ques-
tion to the terrestrial, freshwater, and marine
realms.
FRESHWATER ECOREGIONS
Separate analyses of freshwater and terrestrial
ecoregions were conducted because the distribution
of freshwater biodiversity diverges from terrestrial
patterns. Freshwater ecoregions were based on sev-
eral regional analyses and consultations with re-
gional experts.” Currently, the Global 200 analysis
effectively targets the majority of freshwater prior-
ities. Some targets. however, may change as we near
completion of a global map of freshwater ecoregions
that is based on a standard level of biogeographic
resolution and relevant biomes.
' Victor (1955), Fre x (1971). Zohary (1973), Miva-
— 1975), Yim da a Chinese bn. 78 Map Com-
Department of Conse ation (1087). Noirfalise (1987),
Changchun и of Geography and Chinese Academy
of Sciences (1990), Kurnae 3 Bohn (1994), Krever
et al. (1991), et al. (1995). Ecological Stratifi-
cation Working Group (1995), ed ant et al. (1995). —
(1995). Omernik (1995), Thackway € Cresswell (1995
Mongolian Ministry for — hi the Environment et К)
(1996). European Topic € > on. Nature Conse "n
(2000), Ricketts et al. — К WF/IUC \ (1994,
1997), Bohn & Katenina p Vra Mentes el a
I
^ Hocutt & Wiley (1986), Frest & Johannes (1991),
WCMC (1992), Maxwell et al. (1995). — 1 al.
(1995). Kottelat & Whitten (1996), — et al. (1999).
Abell » al. (2000), Thieme et al. (in pr
Annals of the
206
Missouri Botanical Garden
°1 AQEL Ul рә1ї5 SUOI391000 әш 01 puodsauı02 si9qunu 3y |, *SUOI391099 ООС [FqO[9 эчоеш pue I9Jemysa31] эц, "€ әлпйї 4
uiseq энәх КО
uiseq зәли jews йй
Jejod В aye] eus Шш
¡e109 jeoido.n ¡e3seo) Ш Jayempeay зәли 26187
бицәмап үрә!йод jeyseo> 8 еўәр әли 96187 HEN
Buijjamdn ayesadwe] ¡eyseo) 8 эли 96187 BN
eas pue Jays әдезәйшәј NNI aye] 96107 NI
SUL зәдемцѕә:ч
Да
7 ABR,
[ [тэп
Th.
Volume 89, Number 2
2002
Olson & Dinerstein
The Global 200
207
MARINE ECOREGIONS
Relative to most terrestrial and freshwater ecore-
gions, marine ecological and biogeographic units
are more spatially and temporally dynamic (Sher-
man et al., 1990) and therefore more challenging
to delineate. Marine ecoregions delineated by the
Global 200 are derived from a synthesis of global
and regional spatial schemas, review of the avail-
able literature” and consultations with experts. Kel-
leher et al. (1995), Sherman et al. (1990), Lon-
ghurst (1998), and Bailey (1998) served as the
primary sources for the Global 200. Our base map
does not cover deep water ecosystems (i.e.. pelagic,
abyssal, or hadal) nor are its biogeographic units
as finely resolved as the maps used in the fresh-
water or terrestrial analyses. We believe that sev-
eral forthcoming and detailed analyses of marine
з н. around the world (Callum Roberts.
2001. Allen.
2002) will be useful in testing and improving the
ers. comm. Gerald pers. comm.
accuracy of our results. As in the land-based anal-
yses, the delineation of marine ecoregions is in-
tended to highlight general regions within which
characteristic. animals, plants, ecological interac-
tions, and biophysical processes occur.
BIOGEOGRAPHIC RESOLUTION
The majority of Global 200 regions are composed
of an aggregation of continental-scale ecoregions.
This reflects the coarser level of biogeographic res-
olution applied on a global scale. For example.
whereas 12 terrestrial units were delineated for the
island of New Guinea in the regional analysis (Wik-
2001), only 5 Global 200 units
The ecoregions that were combined
ramanayake et al.,
are recognized.
are adjacent, related by habitat type. and are bio-
geographically similar at a global scale. Appoxi-
mately a third of the ecoregions used in the regional
analyses correspond directly to Global 200 ecore-
The
boundaries of Global 200 ecoregions do not present
gions. ecific location and configuration of
exact target areas for detailed planning. Rather,
Global 200 ecoregions are primarily intended
spotlight regions of exceptional importance for stra-
tegic decision-making.
200
SELECTION CRITERIA FOR THE GLOBAL
Selection of the Global 200 draws heavily from
the results of intensive regional analyses of biodi-
^ Hayden f id SU db IUCN (1988), — (1990).
J et al. 32), Ray & Hayden (1 Kel
al. 5). & Jer
up : Sullivan & Bustamante (
leher et
-
V!
1990). ек (1998
versity conducted over the last several years (Krev-
er et al., 1994; Dinerstein et al., 1995; Olson et al.,
1999; Ricketts et al., 1999; Abell et al.. 2000; Wik-
ramanayake et al., 2001: in press:
Thieme et al
Burgess et al..
in press). Within each biome and
biogeographic realm, the relative importance of
ecoregions was Classified at one of four levels: glob-
ally outstanding, regionally outstanding (e.g.. Neo-
tropics, Atlantic Ocean), bioregionally outstanding
(e.g.. Caribbean), or locally important. The criteria
used to prioritize ecoregions for the Global 200 are
the same as those used for the regional assess-
ments. We chose the set of ecoregions within each
biome that were considered to harbor biodiversity
that was globally outstanding or regionally out-
standing based on the parameters described below.
These parameters were weighted and measured
in the regional analyses as illustrated in Appendix
The weight assigned to the different parameters
varied by biome to better address specific patterns
of biodiversity and ecological dynamics.
SPECIES RICHNESS AND ENDEMISM
Richness values were first corrected for area. We
then divided the range of values for the set of ecore-
gions sharing the same biome and realm into four
categories based on natural breaks. Globally out-
standing ecoregions were compared with those
identified for other realms to ensure consistency. In
general, widely recognized global and regional cen-
ters of richness and endemism were selected for
Global 200 status. The precision of the data varied
considerably as illustrated by richness and ende-
vascular
mism values for ` plants in temperate co-
nifer and tundra biomes (Tables 2. 3).
HIGHER TAXONOMIC UNIQUENESS
The presence of an endemic higher taxon (genus
or family) would contribute more to an ecoregion’s
biotic distinctiveness than would an endemic spe-
cies. Some ecoregions are notable for biotas that
contain unique taxa at higher — levels than
species (Vane-Wright et al., ;
1991; Gaston & Williams,
Williams & Humphries,
moist
Villiams et al.,
=ч Forey et al., 1994;
1994). For example, the
forests of northeastern Australia. northern
New Zealand. and New Caledonia are recognized
as having a number of the most primitive lineages
of conifers and flowering plants in the world (МАЕ
IUCN, 1994-1997)
UNIQUE ECOLOGICAL
PHENOMENA
OR EVOLUTIONARY
200
status because of their extraordinary ecological
Some ecoregions were elevated to Global
208 Annals of the
Missouri Botanical Garden
Table 2. Estimated richness and endemism (expressed as number of species) of native vascular plant species for
temperate coniferous forest pane around the world. Data for ecoregions of the United States and Canada are
derived from the Biota of North America Program databases developed by Kartesz and Meacham (1999). The estimates
for Eurasian ecoregions may be lied: higher than values for the Americas because the former typically en-
compass biogeographic areas that are broader in scope (i.e., they include mixed-conifer and broadleaf forest habitats)
than ecoregions delineated for the Americas (WWE/IUCN, 1994-1997; Mittermeier et al., 1999).
Ecoregion Species richness Endemism
Nearctic
Southeastern Conifer Forests 3095 >201
Sierra Nevada Forests 2373 51-75
Arizona Mountains Forests 2204 76-110
South Central Rockies Forests 1933 51-75
— ae n Forests 1859 111-151
Piney Woods Forests 1729 4-10
North Central * kies Forests 1695 21-50
Colorado Rockies Forests 1626 76-110
Middle Atlantic Coastal Forests 1488 11-20
Okanogan Forests 1355 1-3
Cascade Mountain Leeward Forests 1328 11-20
North Cascades Forests 1325 4-10
Central and Southern Cascades Forests 1296 21-50
Eastern Cascade Forests 1224 21-50
Northern California Coastal Forests 1212 11-20
Blue Mountain Forests 1134 21-50
Wasatch and Uinta Montane Forests 1109 51-75
Central Pacific Coastal Forests 1109 11-20
Puget Lowlands Forests 1100 1-3
Great Basin Montane Forests 1043 21-50
Fraser Plateau and Basin Complex 1012 0
Florida Sand Pine Scrub 951 21-50
Northern British Columbia Mountain Forests 909 0
Northern Transitional Alpine Forests 876 0
Alberta/British Columbia Foothill Forests 740 1-3
Alberta Mountain Forests 660 1-3
Northern Pacific Coastal Forests 615 1-3
Queen Charlotte Islands 459 1-3
Atlantic Coastal Pine Barrens 632 1-3
Neotropics
Valdivian Temperate Rainforests 463 >33
Palearctic
Caucasus Mountains ~6300 ~ 1600
Middle Asia Mountains* ~5500 ~ 1500
Pyrenees ~3500 ~200
Balkan—Rhodope Mountains ~3000 ~900
Alps ~3000 ~350
Carpathians ~2000 ~100
Central China Mixed-Conifer Forests ~ 1900 ?
Eastern Himalayan Temperate Conifer Forests ~1500 ?
* Kopetdag, Tienshan, Pamiro-Alai, Pamir, Dzhungarian Alatau.
phenomena. Ecoregions that contain extensive in- ests of the Amazon (varzea forests) (Goulding,
tact habitats and large vertebrate assemblages were 1980; Goulding et al., 1996). Such phenomena
recognized. Also considered were the long-distance were once widespread but are now rare due to the
migrations of larger terrestrial vertebrates such as prevalence of human disturbance around the world.
caribou or wildebeest, and the tremendous seasonal This is the only situation where we consider global
fish migrations and fish frugivory in the flooded for- — patterns within the context of threats. Otherwise,
Volume 89, Number 2
2002
Olson & Dinerstein
The Global 200
209
Table 3.
comm.).
Actual or estimated vascular plant species richness and endemism (expressed as species — of some
tundra ecoregions or regions based on data from WWF/IUCN (1994). Ricketts et al. (1999), and J. '
Kartesz (pers.
Ecoregion or region Species richness Endemism
Nearctic
Pacific Coastal Mountain Icefields 792 0
Alaska/St. Elias Range Tundre 747 1-10
Interior Yukon/Alaska Alpine ‘Tundra 017 1-10
Brooks/British Range Tundra 193 1-3
Ogilvie/MacKenzie Alpine Tundra 589 1-10
Arctic Foothills Tundra 580 0
Beringia Lowland Tundra 953 0
relic Coastal Tundra 539 1-3
Beringia Upland Tundra 38 1-3
Low Arctic Tundra 197 0
Aleutian Islands ‘Tundra 388 1-10
Middle Arctic Tundra 371 1—3
Torngat Mountain Tundra 286 0
High Arctic Tundra 245 0
David. Highlands Tundra 216 0
Baffin Coastal Tundra 135 0
Palearctic
Chukotsky Peninsula 939 —50
240 5
Taimyr Peninsula
the Global 200 emphasizes biodiversity features
that were in place prior to major human impacts of
natural habitats and species populations.
Both ecological and evolutionary phenomena are
a critical. but widely overlooked. aspect of biodi-
versity conservation. Unusual evolutionary phe-
nomena such as the extraordinary adaptive radia-
tions seen in Hawaiian plants, birds. and insects,
the radiation of Galápagos finches. the radiation of
cichlids in Rift Valley lakes of Africa, also elevated
While evolu-
every
some ecoregions to the Global 200.
tionary or ecological phenomena occur in
ecoregion, we highlight those that are recognized as
exceptional in global comparisons.
GLOBAL RARITY
All ecoregions in globally rare biomes were con-
sidered priorities. We elevated ecoregions to Global
200 status if their biome or major habitat type was
represented in fewer than eight distinct regions
around the world. Examples of rare biomes include
the six Mediterranean woodlands, forests. and
scrub, all of limited area. Temperate rain forest eco-
systems (a major habitat) occur in seven relatively
localized areas around the world. Paramos, or wet
tropical alpine shrublands, occur in only a few ar-
eas of the northern Andes and Central America, a
African and in New
For this criterion, we counted only natu-
few Каз! mountain ranges,
Guinea.
rally occurring rarity, although human-induced rar-
ity is an important condition to assess when devel-
oping conservation strate gies,
INTACTNESS
For ecoregions in the same biome that were as-
sessed at a similar level of biological importance,
we selected the ecoregions that had relatively more
intact habitats and biotas (see conservation. status
below).
REPRESENTATION
Ecoregions were also elevated to Global 200 sta-
tus if they were the best example of their biome
within a realm in situations where no other ecore-
gion had been selected due to its outstanding bio-
diversity. In this selection we emphasized those
r most en-
=
ecoregions that harbored the richest «
demic biotas. or had the most intact natural eco-
systems if biological importance was similar among
candidates,
The Global 200 focuses on biological values as
the critical first step in setting global conservation
priorities. There are many other factors that may be
used in the prioritization process. We purposefully
did not use ecological function, conservation fea-
social, economic, cultural
sibility (i.e. political,
factors), or human utility as discriminators to iden-
Annals of the
Missouri Botanical Garden
tify the Global 200 as these features are either dif-
ficult to measure or are highly fluid. The develop-
ment and implementation of ecoregion strategies
ry
however, require careful attention to ecological
function and non-biological factors.
The recognition of remaining wild animal migra-
tions and other contemporary ecological phenome-
na is the only criterion where human impacts to the
environment are recognized, because areas of ex-
tinguished phenomena are ignored. Otherwise, the
Global 200 emphasizes biodiversity features that
were in place prior to major human impacts on nat-
ural habitats and species populations.
CONSERVATION STATUS OF THE GLOBAL 200
ECOREGIONS
Ecoregions vary greatly not only in their biolog-
ical distinctiveness, but also in their conservation
status. Conservation status represents an estimate
of the ability of an ecoregion to maintain viable
species populations, to sustain ecological process-
es, and to be responsive to short- and long-term
environmental changes. Conservation status assess-
ments of the Global 200 ecoregions were based on
landscape or aquascape-level criteria, such as total
water
quality, and estimates of future threat. From a prac-
habitat loss, the degree of fragmentation,
tical perspective, a measure of conservation status
can dictate the urgency, kinds of conservation ac-
tivities, and level of effort needed among ecoregions
or biomes. Conservation status can also indicate ar-
eas with relatively high opportunity for far-reaching
conservation measures.
e estimated the conservation status of ecore-
gions specifically to enable us to make decisions
about elevating ecoregions when the similarity of
their biodiversity features made discrimination
challenging. Conservation status was also used to
assess broad trends in threats among different re-
gions and biomes. Again, we drew heavily from re-
gional conservation assessments to estimate con-
For the Global 200, we classified
ecoregions into one of three broad categories: erit-
servation status.’
ical/endangered, vulnerable, or relatively stable/
relatively intact over the next 40 years. For terres-
trial ecoregions, the most prominent contributor to
conservation status is habitat loss, followed by the
* IUCN (1991, 1992), Krever et al. (1994), BSP et al.
(1995), Dinerstein et al. (1995), — е! а,
MacKinnon & oe (1996), Bi i bs E
nerstein et al. (19‹ p et al. )7). E. ketts el
al. (1999), Abell et 2 (2000). бош. et en (2000). Con-
servation — (2000). Wikramanayake et
(2001). Burgess et al. ( n press)
size of remaining habitat blocks, degree of frag-
mentation, degree of degradation, and degree of
protection (see Appendix 2). Weightings for factors
varied by biome for freshwater and marine ecore-
gions,
THE GLOBAL 200 ECOREGIONS
We identified 238 ecoregions whose biodiversity
and representation values are outstanding or sig-
nificant on a global scale (Table 1). They represent
the terrestrial, freshwater, and marine realms, and
the 30 biomes nested within these realms. Among
142 (60%) are terrestrial, 53
(22%) are freshwater ecoregions, and 43 (18%) are
the three realms,
marine. Terrestrial ecoregions outnumber those of
the other realms largely because there is more lo-
calized endemism in terrestrial than in marine bi-
otas. Gaps in biogeographic information for fresh-
water and marine biodiversity also account for some
of the variation.
TERRESTRIAL REALM
Tropical and subtropical moist forests
Among the 14 terrestrial biomes, the largest
number of Global 200 ecoregions falls within the
tropical and subtropical moist forests biome (50
ecoregions or 35% of all terrestrial ecoregions) (Ta-
ble 1). The high number of ecoregions reflects the
biological richness and complexity of tropical moist
forests. Although there are more tropical moist for-
est ecoregions in the Indo-Malayan Biogeographic
realm (17) than in the Neotropics (12), this is partly
due to the archipelagic distributions of Asian trop-
ical moist forests and their characteristic biotas
(Whitmore, 1986, 1990; Whitten et al. 1987a.
1987b, 1996; Wikramanayake et al., 2001). Four
of the Asian tropical moist forests are small island
systems, and the original extent of all of the Asian
ecoregions fits easily within the area covered by
western Amazonian moist forests.
The most diverse terrestrial ecoregions occur in
the Western Arc forests of the Amazon Basin, with
close rivals in the Atlantic Forest ecoregion of Bra-
zil, the Chocó-Darién ecoregion of northwestern
South America, Sumatra, and Peninsular Malaysia
and northern Borneo forest ecoregions. The mon-
tane forest biotas of the Northern Andes are re-
markable for their globally high rates of beta-di-
versity and extraordinary local endemism (Terborgh
& Winter, 1983; ICBP, 1992; Hamilton et al., 1995;
Wege & Long, 1995; WWF/IUCN, 1994—1997).
The forests of the Guayanan region and Cuba are
known for their pronounced endemism and unusual
Volume 89, Number 2
2002
Olson & Dinerstein
The Global 200
biogeographic relationships (Whitmore & Prance,
1987: Borhidi. 1991; Dinerstein et al., 1995; Stey-
ermark et al.. 1995; Hedges, 1996). The forests of
the Greater Antilles also are notable for a number
of relict mammals, such as solenodons and hutias.
The Congolian coastal forests are likely the most
diverse in the Afrotropics, although diversity infor-
mation is scarce for several ecoregions in the cen-
tral Congo Basin (Oates, 1996; Kingdon, 1997:
Burgess et al., in press). The Guinean moist forests
support many species not found in the Central Af-
rican region (IUCN/UNEP. 1986a: IUCN. 1990:
Martin. 1991: IUCN., 1992: Mittermeier et al..
1999). The Albertine Rift montane forests are ex-
tremely rich for some taxa, such as birds. and have
a high degree of endemism (Collar & Stuart. 1988:
Kingdon, 1989: WWE/IUCN, 1994). The distinc-
tiveness of the Eastern Are montane and East Af-
rican coastal forests is attributable to their great age
and isolation (Hamilton € Bensted-Smith, 1989:
Lovett & Wasser, 1993: Hamilton et al., 1995; Bur-
gess et al., in press). Madagascar forests and shrub-
lands are also highly distinctive on global scales.
especially at higher taxonomic levels (Nicoll &
Langrand, 1989; Preston-Mafham, 1991: WWE/
IUCN, 1994). Tropical moist forests of New Guinea
are highly distinctive (Brooks. 1987: Flannery,
1990. 1994: WWF/IUCN, 1994; Mittermeier et al..
1996: Wikramanayake et al., 2001). although Aus-
tralian moist forests do share many affinities with
New Guinea. The long-isolated forests of New Cal-
edonia are exceptionally unusual, with so many en-
demic and relict higher taxa and species that the
island is considered the “Madagascar of the Pacific.”
The forests of Sulawesi are noted for their region-
ally high degree of endemism in a range of taxa, a
phenomenon also seen in the Philippine moist for-
Lesser Sundas semi-evergreen for-
(IUCN/UNEP, 1986b; IUCN, 1991: ICBP.
2: Stattersfield et al.. 1998; Wikramanayake et
al.. 2001). The Western Ghats and southwestern Sri
Lankan moist forests are distinctive due to their
ests and in the
isolation and stability of conditions over millions of
years. Tropical moist forests on oceanic islands are
often highly distinctive due to high rates of ende-
mism, extraordinary radiations of taxa and adaptive
radiation, and relictual or unique higher taxa (Dahl.
1986: Mitchell, 1989; Johnson & Stattersfield.
1990; Flannery, 1994: WW F/IUCN, 1994; Wagner
& Funk. 1995).
Tropical and. subtropical dry forests
The most diverse dry forests in the world occur
in southern. Mexico and in the Bolivian lowlands
(Gentry, 1993; Parker et al.. 1993: Bullock et al.,
1996). The dry forests of the Pacific Coast of north-
western South America support a wealth of unique
species due to their isolation (Parker & Carr. 1992:
WWF/IUCN, 1994: Bullock et al.. 1996).
tropical forests of Maputaland-Pondoland in south-
eastern Africa support many
endemies (Cowling € Hilton-Taylor, 1994: WM F/
IUCN, 1994). The dry forests of central India and
Indochina are notable
tebrate faunas (Corbett & Hill.
1995). Dry forests of Madagascar and New
donia are globally distinctive because of their high
The sub-
are diverse and
for their diverse large ver-
1992: Stewart-Cox,
Cale-
taxa and extreme endemism
1987: Preston-Mafham. 1991:
Wikramanavake et al.. 2001).
number of relictual
(IL CN/UNEP/WW EF.
WWE/IL CN, 1994;
Tropical and subtropical coniferous forests
the world’s richest and most
complex subtropical coniferous forests (Perry.
1991: Peterson et al., 1993: Ramamoorthy al..
1993; WWE/IUCN, 1994).
the Antilles contain many endemics am
relictual taxa (Borhidi, 1991). Subtropical conifer
forests of Indochina are incorporated into the dry
Mexico harbors
The conifer forests of
Greater
and moist forests of the region.
Temperate broadleaf and mixed forests
Temperate broadleaf and mixed forests are rich-
est in central China and eastern North America,
with other globally distinctive ecoregions occurring
in the Caucasus, the Himalayas, southern Europe.
and the Russian Far East (Table 2) (Zhao et al.,
1990: Martin et al., 1993: Oosterbroek. 1901:
WWF/IUCN, 1994; Mackinnon € Hicks. 1996:
Ricketts et al.. 1909),
Temperate coniferous forests
Temperate rain forests only occur in seven re-
gions around the world—the Pacific Northwest. the
Validivian forests of southwestern South America,
rain forests of New Zealand and Tasmania. the
in
the 1
Northeastern. Atlantic (small.
Ireland. Scotland. and Iceland).
pan. and those of the eastern Black Sea (Kellogg
t al.. 1992: WWE/UCA. 1994). Forest commu-
nities dominated by huge trees (e.g.. giant sequoia.
Sequoiadendron gigantea (Lindl.) J. Buchholz: red-
wood, Sequoia sempervirens (D. Don) Endl.: moun-
tain ash. Eucalyptus regnans F. Muell.) are unusual
isolated. pockets
southwestern Ja-
ecological phenomena that are found only in west-
ern North America, America,
in the Australasian region in such areas as
southwestern South
and
Annals of the
Missouri Botanical Garden
southeastern Australia and northern New Zealand.
The Klamath-Siskiyou ecoregion of western North
America harbors diverse and unusual assemblages
and displays notable endemism for a number of
plant and animal taxa.
Chile are notable for their diversity of tree genera,
many of which are monotypic and have Gondwan-
These long-isolated forests have
aland origins.
many other unusual taxa and unique communities.
Boreal forests and taiga
Low species richness and endemism are char-
acteristic of circumboreal and cireumpolar ecore-
gions (USSR Academy of Sciences, 1988), thus the
presence of intact ecological phenomena denoted
outstanding ecoregions. Large-scale migrations of
caribou, or reindeer (Rangifer tarandus), and intact
predator assemblages can still be found in some
regions. For example, the Northern Cordillera bo-
real forests of Canada have been called the Ser-
engeti of the Far North due to their abundance and
‹
-—
м
diversity of large vertebrates (Ricketts et al.,
Extensive tracts of boreal forest and taiga still exist
in the northern Nearctic and Palearctic, the largest
expanses being in central and eastern Russia
(Stewart, 1992; Krever et al., 1994).
also enjoys relatively unaltered natural disturbance
This biome
regimes, an increasingly rare situation in other bi-
omes.
Tropical and subtropical grasslands, savannas, and
shrublands
In many parts of the tropics large mammal fau-
nas have evolved to take advantage of the produc-
tive grasses and browse typical of this biome. These
large mammal faunas are richest in African savan-
nas and grasslands. Presently the most intact as-
semblages occur in East African acacia savannas
and Zambezian savannas comprised of mosaics of
miombo, mopane, and other habitats (McClanahan
& Young, 1996). Large-scale migration of tropical
savanna herbivores, such as wildebeest (Conno-
chaetes taurinus) and zebra (Equus zebra), are con-
tinuing to decline through habitat alteration and
hunting. Only in East Africa, the central Zambezian
region, and in the Sudd region (Uganda kob or Ko-
bus kob) do sizable migrations still persist. Many of
the extraordinary migrations of the Guinean and Sa-
helian savannas have disappeared. Sahelian ecore-
gions support a large number of endemic rodent
species, while the Somalian bushland and thickets
harbor a concentration of endemic mammals, from
rodents to antelopes. Both the Cerrado and the Lla-
nos are noted for complexity of habitats and the
The Valdivian forests of
unusually high levels of endemism and beta diver-
The tropical
savannas of northern Australia and southern New
sity in plants for tropical savannas.
Guinea support distinctive communities with sev-
eral pockets of endemism for a range of taxa (Stat-
tersfield et al., 1998).
Temperate grasslands, savannas, and shrublands
The vast expanses of grass in North America and
Eurasia once sustained vast migrations of large ver-
lebrates such as buffalo (Bison bison) and saiga
(Saiga tatarica). Such extraordinary phenomena
now occur only in isolated pockets, such as on the
1994; Hilbig, 1995;
The extraordinary floral communities
Daurian Steppe (Krever et al.,
1996).
of the Eurasian steppes and the North American
Finch,
Great Plains have been largely extirpated through
ЗОО different
plant species can occur on a few hectares of North
conversion to agriculture. Nearly
American tallgrass prairie. The Patagonian steppe
and grasslands are notable for endemic higher taxa
for mammals.
Flooded grasslands and savannas
Some globally outstanding flooded savannas and
grasslands occur in the Everglades, Pantanal, Sa-
helian flooded savannas, Zambezian flooded savan-
nas (including the Okavango Delta), and the Sudd.
The Everglades are the world's largest rain-fed
flooded grassland on a limestone substrate. The
looded savannas and erre selected are gen-
erally the largest complexes in each region. Anoth-
er extraordinary inland delta, the Mamberamo Riv-
er inland delta, is captured. within the montane
forests of the New Guinea ecoregion,
Montane grasslands and shrublands
The paramos of the northern Andes are the most
extensive examples of this biome. Paramo ecosys-
te ms occur in only a few other localities in the trop-
. The heathlands and moorlands of East Africa
(e. Mt. Kilimanjaro, Mt. Kenya, Rwenzori Mts.,
Ethiopian Highlands), Mt. Kinabalu of Borneo, and
the Central Range of New Guinea are all limited in
extent, extremely isolated, and support highly en-
demic plants and animals. A characteristic feature
of many tropical paramos is the presence of large
rosette plants from a variety of plant genera, such
as Lobelia (Africa), Puya (South America), Cyathea
(New Guinea), and Argyroxiphium (Hawaii)—these
plant forms can reach elevations of 4500-4600 m
above sea level. Drier, yet distinctive, subtropical
montane grasslands, savannas, and woodlands in-
Volume 89, Number 2
2002
Olson & Dinerstein 213
The Global 200
clude the Ethiopian Highlands, the Zambezian
montane grasslands and woodlands, and the mon-
tane habitats of southeastern Africa (Werger, 1978:
White, 1983; Huntley, 1989, 1994; Timberlake &
Müller, 1994; WWF/IUCN, 1994). The montane
erasslands of the Tibetan Plateau still support rel-
Pan-
—
atively intact migrations of Tibetan antelope
tholops hodgsoni) and kiang, the Tibetan wild ass
(Equus hemionus). The puna grasslands of the high
Andes support over 30 species of endemic rodents
(45 total species).
Tundra
Tundra ecoregions were selected. primarily. be-
cause of extraordinary seasonal concentrations of
breeding waterfowl and shorebirds. and caribou
(Stewart, 1992: Krever et al., 1994; Ricketts et al..
1999). Relatively intact. tundra ecoregions were
chosen, wherever possible. The Chukotsky tundra
ecoregion is unusual with nearly 50 endemic plant
Knystautas, 1987; USSR Academy of Sci-
1988: WWE/IUCN, 1994).
species
—
ences,
Deserts and xeric shrublands
The
support the world’s richest desert floras (Cowling &
1994: Maggs et al., 1994: WWF/
while the Chihuahuan Desert and
Namib—Karoo deserts of southwestern Africa
Hilton-Taylor,
IUCN, 1994),
central Mexican deserts are a close second and are
the richest Neotropical deserts (Cowling et al..
1989: Hernandez & 1995; Ricketts et
al.. 1999). Australian deserts support the richest
reptile faunas. The Carnavon Xeric Scrub of west-
Barcenas.
ern Australia is a regional center of endemism for
a range of taxa. Unusual desert communities dom-
inated by giant columnar cacti occur in the Sonoran
and Baja Deserts of North America (Brown, 1991).
while the spiny thickets of southwestern Madagas-
car are globally unique in terms of structure and
taxa. Some Baja California communities are par-
tially convergent in structure with the Madagascar
thickets.
South America (including the adjacent transition
The Atacama Desert ecoregion of western
area of the Monte/Puna/Yungas) and the Horn of
Africa deserts were recognized as some of the more
outstanding regional centers of richness and en-
demism. The Central Asian deserts. while not as
rich as Afrotropical or Neotropical deserts, are rep-
resentative of the region’s deserts with diverse rep-
tile and mammal faunas.
Mediterranean forests. woodlands, and scrub
АП five Mediterranean-climate ecoregions are
highly distinctive, collectively harboring 20 percent
of the Earth's plant species (Cody, 1980; Kalin Ar-
1995; Picker & Samways, 1995). Phy-
togeographers consider the Fynbos as а separate
floral kingdom because 68% of the 8600 vascular
plant species crowded into its 90.000 km? are en-
royo el al.,
demie and highly distinctive at several taxonomic
1989, 1996; Cowling & Hil-
ton-Taylor, 1994). In terms of species densities, this
levels (Cowling et al.,
is equivalent to about 40 percent of the plant spe-
cies of the United States and Canada combined,
found within an area the size of the state of Indiana
(N. Myers,
west Australia shrublands have floras that are sig-
pers. comm.). The Fynbos and South-
nificantly more diverse than the other ecoregions,
although any Mediterranean shrubland is still rieh
in species and endemics relative to other non-forest
ecoregions (Cowling et al, 1996: Oosterbroek,
199
Mangroves
The diversity of mangroves in the Indo-West Pa-
cific (WP) region is much greater than those of the
Atlantic—Caribbean—East Pacific (ACEP) region—
the former supporting 17 genera and 40-42 species
of true mangroves and the latter having only 4 gen-
era and 7 (MaeNae, 1968: Lacerda, 1993;
Olson et al... 1996; Spalding et al., 1997; Rickleffs
& Latham, 1993). A single site in the ACEP typi-
cally contains 3 or 4 true mangrove species, while
species
30 species have been recorded from one locality in
the IWP region (Ricklefs & 1993
grove forests on the western coast of НЫЕ
Latham, lan-
support a number of endemic bird species that are
endangered. The mangrove swamps and forests of
the Indo-Malayan and Australasian realms are the
world's most extensive. South and Southeast Asia
12% of the total area of the world’s
1997).
are the largest contiguous mangrove forest in the
alone contain
mangroves (Spalding et al.. The Sundarbans
world. The vast floodplains of New Guinea also sup-
port extensive mangrove swamps unrivaled. else-
where in the world.
and terrestrial
If all of the marine, freshwater,
species that occur in mangroves are considered,
these seemingly simple forests can be considered
as one of the more diverse ecosystems in the trop-
ics. Mangroves are keystone habitats in the sense
that they have an inordinately strong influence on
species populations and ecosystems well beyond
their limited area. In addition to providing habitat
and resources to a wide range of species, mangrove
forests and swamps also protect inland habitats and
shorelines from damage by damping storm waves
and tidal action. Mangroves filter silt and pollutants
214
Annals of the
ende Botanical Garden
from terrestrial runoff that would otherwise damage
seagrass beds and coral reefs
FRESHWATER REALM
Large rivers
Faunas adapted to high-flow regimes of large riv-
ers are uncommon and best developed in the Yang-
tze, Colorado, and lower Congo Hivers. A relatively
small area of rapids in the latter region supports 22
endemic species of fish that are rapid specialists
(Lowe-McConnell, 1987).
raná, and Amazon—Orinoco Rivers harbor the four
The Mekong, Congo, Pa-
Р, B
great large tropical river fish faunas (Mori, 1936;
Roberts, 1975; Hocutt & Wiley, 1986; Lowe-
McConnell, 1987; Kottelat & Whitten, 1996). The
waters of the Lower Yangtze and Mississippi Rivers
contain outstanding examples of large-river fishes,
amphibians, reptiles, and invertebrates, including
relicts and many endemics (Abell et al.. 2000).
Large river headwaters
Species, assemblages, and processes in head-
water areas are distinct from those of thei мг larger
mainstems. The Mississippi Piedmont, Guayanan
highlands, Upper Amazon, Upper Parana, Brazilian
Shield, and Congo Basin Piedmont harbor a tre-
mendous array of species, including numerous en-
demics adapted to life in these waters. In turn,
these river systems ultimately feed a number of the
world’s largest and richest rivers (Hocutt & Wiley,
1986; Kottelat & Whitten, 1996;
press). The most diverse vertebrate assemblages on
Thieme et al., in
Earth occur in freshwater communities of the Am-
azon and the Orinoco River basins. Over 3000 spe-
cies of fish are estimated to occur in the Amazon
Basin alone (Goulding, 1980).
Large river deltas
Delta complexes of several large temperate and
polar rivers are identified, including the Mesopo-
tamian, Volga, and Lena River deltas. The Niger
River delta, the most extensive river delta in Africa,
is characterized by high species richness (Wetlands
International and The World Bank, 1996;
et al., in press). The extensive deltas of the Orinoco
Thieme
and Amazon Rivers are encompassed in their re-
spective large-river ecoregions (see above).
Small river basins
The Mississippi River embayment, the Mobile
River basin, and numerous coastal streams and riv-
ers of southeastern North America together support
one of the Earth’s richest temperate freshwater bi-
otas (Hocutt & Wiley, 1986; Hackney et al., 1992;
Abell et al., 2000). The headwater streams and riv-
ers of the Yangtze River in central China are also
extremely diverse (recognized as a large river bi-
ome in this analysis) (Mori, 1936; Nichols, 1943;
Taki, 1975). Secondary centers of temperate diver-
sity occur in the rivers and streams of southeastern
North America, the western coast of North America,
and the Russian Far East (Zhadin & Gerd, 1961;
ee et al., 1980; Hocutt & Wiley, 1986; Groom-
bridge & Jenkins, 1998; Abell et al., 2000). Several
freshwater biotas on islands are highly distinctive,
including those of Madagascar, New Guinea, the
the Greater Antilles, Sri Lanka,
New Caledonia (IUCN/UNEP/WWE, 1987;
Zakaria-Ismail, 1987, 1994; Allen, 1991; Preston-
Mafham, 1991; Oberdorff et al.. 1995). The South-
west Australian Rivers and streams ecoregion is a
Greater Sundas.
and
center of endemism, while also harboring a number
of primitive higher taxa and several species with
highly unusual freshwater life histories (McDowall,
1996; State of the Environment Advisory Council,
1996). Rivers and streams along the Gulf of Guinea
harbor some of the richest and most endemic riv-
erine freshwater biotas in Africa (Kingdon, 1989;
1992; Lévéque, 1997;
al.. in press). The Salween River of Southeast Asia
Lévêque et al., Thieme et
is recognized for its rich and endemic freshwater
fish fauna (WCMC, 1992). The rivers and streams
` New Guinea, including the inland delta of the
Mamberamo River of New Guinea, support a large
number of unusual and endemic species and higher
taxa.
Large lakes
The Global 200 also identifies the most outstand-
ing examples of diverse and endemic freshwater
faunas in large lakes found in temperate and trop-
ical regions, many displaying extraordinary species
flocks and adaptive radiations in fish taxa. Some
particularly notable lake biotas include those of the
African Rift
Lake Baikal,
high-altitude lakes of the Andes, and the highland
akes of Mexico (Myers, 1960; Roberts, 1975; Ho-
cutt & Wiley, 1986; Allen, 1991; Stiassny et al.,
1992; WCMC, 1992; Nagelkerke et al., 1995; Kot-
telat & Whitten, 1996; Olson et al., 1999; Thieme
et al.,
Lakes and Lake Tana in Ethiopia,
Lake Biwa of southern Japan, the
in press).
Small lakes
Similarly, a number of smaller lakes around the
world host extraordinary expressions of freshwater
Volume 89, Number 2
2002
Olson & Dinerstein 215
The Global 200
Lake Kutubu and Lake Sentani of New
Lakes
Highland Lakes, the Cameroon Crater Lakes, Lake
biodiversity.
Guinea, Yunnan and Streams, Mexican
Lanao of the Philippines, Lake Inle Myanmar
(Burma). and the Central Sulawesi Lakes have been
selected for their globally outstanding biodiversity
features.
Yeric basins
Ephemeral streams, rivers, and lakes. and per-
manent springs characterize ecoregions in this bi-
ome. Low richness and high endemism in fish and
invertebrates (e.g.. molluses) is typical of the Chi-
Anatolian, and Central Australian fresh-
1980: Balik.
The Cuatro Ciénegas
huahuan,
water ecoregions (Hocutt & Wiley,
1995; Abell et al., 2000).
spring and pool complex in the Chihuahuan Desert
is globally unique in its high richness, extreme en-
demism. and unusual evolutionary adaptations
(Contreras-Balderas, 1978; Hocutt & Wiley. 1986).
Freshwater habitats in the Anatolian region of Tur-
key support many endemic species (Balik. 1995).
MARINE REALM
The distribution of marine biodiversity varies
widely throughout ocean basins (Briggs. 1974: El-
der & Pernetta. 1991: Angel, 1992, 1993: Clarke.
1992: Kendall & Aschan. 1993; Kelleher et al..
1995; Groombridge & Jenkins, 1996;
al.. 1997). The abundance and diversity of most
taxa tend to be highest near continental and island
Ormond et
margins that are less than 2000 m deep (Кау. 1991:
Johannes & Hatcher, 1986; Gray, 1997). These ar-
eas experience nutrient enrichment from upwelling
processes and terrestrial runoff (Ray. 1988: Norse.
1995). Areas where significant upwelling occurs are
often. extraordinarily productive in tropical. tem-
perate, and polar regions. Within biomes, species
richness and endemism also vary enormously
around the globe.
Current biogeographic data suggest that species
endemism tends to be less pronounced in marine
ecosystems than in terrestrial or freshwater ecore-
gions, but several regional centers of endemism are
recognized, including the southern coast of Austra-
lia, New
lands. the northern coast of South America, the Yel-
Каз! Hed Sea. the
Mediterranean Sea, the Sea of Cortez, the
Caledonia. Lord Howe and Norfolk Is-
low and China Seas, the
Greal
Barrier Reef, and tropical Pacific Islands such as
Hawaii. the Marquesas, the Tuamotus and Socie-
ties. and Easter Island (Robbins. 1991: Lieske &
Myers, 1996; Vernon, 1995; Groombridge & Jen-
kins, 1996). In general, marine ecoregions associ-
ated with isolated islands and enclosed seas tend
to display pronounced endemism (Kelleher et al..
1995; Groombridge € Jenkins, 1996).
We categorized the marine realm into 10 biomes.
Pelagic (trades and westerlies). abyssal. and hadal
biomes, however, were not assessed for the Global
200 marine analysis because of the large scale of
these units compared to other Global 200 ecore-
gions. the lack of consensus on their classification.
and the limited biodiversity information for these
1991: Grassle. 1991:
Large biogeographic
ecosystems (see Gage « Tyler,
Grassle & Maciolek, 1992).
units have been identified for pelagic and abyssal
biotas (e.g.. Brinton, 1962: Angel. 1993: Longhurst.
1998: Pierrot-Bults, 1997: Vinogradova. 1997). but
their scale is several orders of magnitude greater
Global 200
units may be biogeographically and dynamically
than most ecoregions. These larger
appropriate for open ocean environments. The vast
size and dynamic nature of these biomes precluded
delineating biogeographic subunits at an appropri-
ate level of resolution for the Global 200. Pelagic
species are noted for widespread distributions.
while the few ocean trench surveys that are avail-
able suggest many species are endemic to single
trenches. The paucity of species data for these eco-
systems also reduces our confidence to undertake
comparative analyses.
Polar
The Weddell Sea and Peninsular Antarctica were
identified as the most productive and diverse ecore-
gions of the Antarctic large marine ecosystem. The
Bering. Beaufort, and Chukchi Seas and Barents—
Kara Seas ecoregions are arguably the two most
diverse and productive Arctic marine ecosystems
(USSR Academy of Sciences. 1988: Reeves «
1994). Marine ecosystems near
southern Greenland require further evaluation,
Leatherwood,
Temperate shelf and seas
Some of the most productive marine ecosystems
occur in the Grand Banks and New Zealand plus
The South
coastal waters are remarkable for unusually high
the Patagonia ecoregions. Australian
invertebrates and
=
levels of endemism in some
groups of fish, in addition to the diverse marine
mammal assemblage found there. Two of the world’s
argest temperate estuaries, the Chesapeake and
Atlantic Shelf
are elevated to the Global 200 due to their size.
Delaware Bays. and the Northeast
productivity, and habitat diversity. Some of the most
distinctive enclosed temperate seas. the Mediter-
216
Annals of the
Missouri Botanical Garden
ranean Sea and the Yellow—East China Seas, are
recognized in the Global 200.
Temperate upwelling
Highly productive and diverse coastal upwelling
areas occur along the West Coast of North America
where the California Current moves southward.
Along the southwest coast of Africa the Benguela
Current exhibits similar dynamics.
Tropical upwelling
The Humboldt Current along the West Coast of
South America and the Canary Current along the
West Coast of Africa bring rich nutrients to the sea
surface where they support highly productive ma-
rine systems. Important tropical upwelling and cur-
rent areas also occur in the Panama Bight ecore-
gions.
Tropical coral
Southeast Asian seas support more than 450 spe-
cies of hard (scleractinian) corals, the western In-
dian Ocean around 200, and the Caribbean only 50
species (Vernon, 1995). Variation in reef fish and
non-coral invertebrate diversity follows a similar
biogeographic pattern (McAllister et al.. 1994;
Lieske & Myers, 1996). Overall, the coral reef com-
munities of the central Indo-Pacific seas are the
most diverse in the world, with the Sulu, Sulawesi,
Banda, and Coral Sea ecoregions being the most
diverse on Earth (Vernon, 1995; Lieske & Myers,
1996). The largest barrier reef in the world is the
Great Barrier Reef. Other world-class barrier reefs
include the barrier reefs of New Caledonia, the Me-
soamerican and Bahamian barrier reefs, and the
large barrier reefs of Fiji. The largest coral atoll
complexes occur in the Maldive-Lakshadweep
ecoregion of the central Indian Ocean and in the
Tuamotus of the central Pacific.
CONSERVATION STATUS OF ECOREGIONS
Among all terrestrial Global 200 ecoregions (142
in total), 75 ecoregions (53%) are considered crit-
ical or endangered, 39 ecoregions (27%) vulnera-
ble, and 28 ecoregions (20%) relatively stable or
intact (Table 1). Terrestrial ecoregion boundaries do
not reflect the extensive habitat loss, fragmentation,
and degradation that have occurred in many of the
terrestrial ecoregions. In ecoregions that have been
dramatically altered, characteristic species and
communities survive only in the few remaining
small blocks of habitat (e.g.. Collar & Stuart, 1988;
Dinerstein et al., 1995). Among the terrestrial bi-
omes, ecoregions falling within the tropical and
subtropical dry broadleaf forests, temperate grass-
lands, Mediterranean shrublands, and temperate
broadleaf and mixed forests are the most threat-
ened. Virtually all biotas on small islands are vul-
nerable or critical/endangered due, in large part, to
their limited habitat area and extreme sensitivity to
anthropogenic disturbance — alien species (Ra-
1988; Wilson, 1988, 1992; WCMC, 1992; Su-
jatnika et al., 1995; Brooks et al., 1997; Reaka-
Kudla et al., 1997). Island ecoregions are projected
to experience a wave of extinctions over the next
ven,
two decades given the fragility of island ecosys-
tems, the sensitivity and endemicity of island spe-
cies, and the severe threats native island biotas
face worldwide. Mangrove habitats are threatened
worldwide from a range of threats including clear-
ing and channelization for shrimp ponds, aquacul-
ture, and agriculture, the extraction of timber and
fuelwood, pollution, and habitat loss due to urban
and industrial expansion.
Assessment of conservation status for freshwater
North America and South America
was based on existing regional analyses (Abell et
al., 2000; Olson et al., 1999). In Africa and Europe,
analyses currently under n
press) provided the basis for rankings presented
here. In areas where no regional assessment has
ecoregions
way (Thieme et al., i
been undertaken, review of relevant literature fa-
cilitated decisions on the levels of threat faced by
native biotas. Worldwide, freshwater organisms rep-
resent a disproportionate number of endangered
species; thus, it is not surprising that so many
freshwater ecoregions received a critical rating in
the assessment. In particular, seasonally flooded
forests, cataracts, and freshwater communities in
xeric areas, are endangered worldwide (Goulding et
al., 1996; Abell et al., 2000; Olson et al.. 1999).
Moreover, most temperate freshwater biotas are
threatened by invasion of exotics, pollution, dams,
and habitat degradation. Among the 53 freshwater
ecoregions 31 (58%) were deemed to be critical or
endangered, 10 (19%) were assessed as vulnerable,
and only 12 (23%) were assessed as relatively sta-
=
=
The individual status of marine ecoregions was
estimated through review of the literature and con-
sultations with regional specialists. Twelve marine
ecoregions (29%) were considered relatively stable
or intact, while another 12 (29%) were considered
critical or endangered. In marine biomes, upwelling
areas are heavily overfished, enclosed seas are de-
graded, and coral reefs and mangroves are severely
affected by habitat destruction, degradation, and
overfishing around the world (Sherman et al., 1990;
Volume 89, Number 2
2002
Olson & Dinerstein 217
The Global 200
Suchanek, 1994; Kelleher et al.. 1995; Bryant et
al.. 1995: Olson et al.. 1996; Ormond et al.. 1997).
Increasingly rising sea surface temperatures from
global warming may endanger all coral reef ecore-
gions within several decades.
DEGREE OF OVERLAP OF TERRESTRIAL,
FRESHWATER, AND MARINE GLOBAL 200
ECOREGIONS
The linkages among terrestrial, freshwater, and
marine conservation are often overlooked. Among
the Global 200, 33 (23%) of the ТАЗ terrestrial
ecoregions overlap extensively with freshwater
ecoregions (i.e., more than 50% of the original ex-
tent of the terrestrial ecoregion is covered by a
freshwater unit). Thirty-four (23%) of the terrestrial
ecoregions share at least 50% of their coastline
with a marine ecoregion. Ten (6%) of the terrestrial
ecoregions do both, overlapping extensively with a
freshwater ecoregion and sharing at least 50% of
their coastline with a marine ecoregion. The terres-
trial ecoregions of this third group are the Mada-
gascar dry forests, Congolian coastal forests, Great-
er Antilles moist forests, Pacific temperate. rain
forests of North America, Queensland tropical
moist forests, southeastern. Australia. Aucalyptus—
Acacia forests, New Caledonia moist forests, New
Caledonia dry forests, New Guinea lowland forests,
Sulawesi moist forests, Philippine moist forests,
Northeast Borneo/Palawan moist forests, and Rus-
sian Far East temperate forests. Carefully designed
conservation activities in these 13 units could ul-
timately affect 39 ecoregions.
THE GLOBAL 200 as A CONSERVATION TOOL
The Global 200 is based on the best available
information and biological insights. As new inter-
pretations of biogeography and better information
on the distribution of species and phenomena be-
come available, we expect to periodically revise the
Global 200. The present list and map incorporate
a number of changes from an earlier version (Olson
& Dinerstein, 1998). For example, the highly un-
usual freshwater biota of southwestern Australia is
now recognized, and the terrestrial ecoregions of
the Amazon Basin have undergone major revisions
based on a recent biogeographic analysis by Silva
(1998).
EXPANDING CONSERVATION GOALS
The Global 200 goes beyond the conservation
targets of other prominent global priority-setting ef-
forts by explicitly incorporating representation
guidelines for biomes within realms. Biological
phenomena are also important criteria used in ils
selection protocol. The Global 200 also emphasizes
freshwater and marine. biodiversity. The Hotspots
analysis (Mittermeier et al.. 1999; Myers et al.,
2000), for example, mostly targets very large and
threatened terrestrial regions with concentrations of
range-restricted (locally endemic) species. The
Hotspots are largely nested within the Global 200
—
> 90% congruence) because both analyses em-
phasize exceptional levels of endemism for species
and higher taxa. The Global 200 can complement
hotspot analyses by corroborating the vast majority
of their priority areas and, in some cases, by pro-
viding a finer resolution of the variation of biodi-
versity features within important regions. For ex-
ample, the Madagascar Hotspot identified by Myers
et al. (2000) corresponds to five separate Global
200 ecoregions and the Indo-Burma Hotspot over-
laps with 14 Global 200 terrestrial and freshwater
ecoregions. The Global 200 also encompasses dis-
tinct freshwater and marine hotspots and warm-
spots, as well as ecoregions important for their ex-
traordinary ecological or evolutionary phenomena
and their representation value. Endemic Bird Areas
of the World highlights concentrations of bird spe-
cies with restricted ranges (Stattersfield et al.,
1998). Like hotspots, the majority of the Endemic
Bird Areas are nested within the Global 200. Both
Tropical Forest. Wilderness Areas (Mittermeier el
al.. 1999) and Frontier Forests (Bryant et al., 1997)
map larger landscapes of relatively undisturbed
natural forests around the world. Although the
Global 200 does not specifically employ forest wil-
derness as a discriminator, again there is extensive
overlap with these wilderness areas because such
areas often harbor rich assemblages of species and
endemics, and unusual phenomena such as intact
predator-prey systems.
OTHER CONSERVATION TARGETS
Other conservation targets, such as species of
special concern, keystone species, habitats, and
phenomena, large-scale ecological phenomena
(e.g. bird, butterfly, caribou, cetacean, sea turtle
migrations), wilderness areas, ameliorating climate
change impacts, reducing toxins, and maintaining
ecosystems with low impacts from alien species are
the Global 200.
Again, effective conservation within priority ecore-
also not directly addressed by
gions and coordinated efforts among ecoregions will
help achieve conservation goals for these targets.
218
Annals of the
Missouri Botanical Garden
AN AMBITIOUS BLUEPRINT FOR GLOBAL
CONSERVATION
One tactical concern of the Global 200 is that it
is ambitious, and that by focusing on 238 ecore-
gions rather than on a handful of conservation
units, we run the risk of placing less emphasis on
the most diverse and distinct ecoregions. In re-
sponse, we maintain that the broad geographic
reach of the Global 200 makes almost every nation
on Earth a stakeholder in a global conservation
strategy. From the global scale to regional and na-
tional-level conservation strategies, the Global 200
lends weight to shared priorities and provides a
global perspective for lobbying efforts by local con-
servation groups. The Global 200 also can help ma-
jor development agencies better recognize and mit-
igate the effects of projects that result in land use
change, or forego development activities in partic-
ularly important and sensitive ecoregions.
The targets of the Global 200—representation,
outstanding ecoregions, and ecological phenome-
na
are all essential elements of a global conser-
vation strategy. The conservation community should
not shrink from this ambitious but necessary agen-
da. The widespread destruction of the Earth's bio-
diversity occurring today must be matched by a re-
sponse at least an order of magnitude greater than
currently exists. The Global 200 provides a nec-
essarily ambitious template for a global conserva-
tion strategy.
Literature Cited
Abell, R., D. M. Olson, E. Dinerstein, P. Hurley ‚1. Т.
Diggs, W. Eichbaum, S. Walters Wettengel, T. All-
nutt, C. Loucks & P. Hedao. 2000. Freshwater —
A C conservation Assessment,
Island Puer АЕ . D.
Alle "n, G. 1991. Field Guide 1 to He pede Fishes
of New ae Publication No. 9. Christensen. Re-
searc А pros Madang, Papua Ne W Guin
Angel, 1992. Managing biodiversity in he oceans.
Pp. У, — їп М. М. A. Peterson (editor), Diversity of
Ocean Life. Center for Strategic and e Stud-
ies, ees DA
‚ 19 lives вну of the pelagic ocean. Conser-
vation B a T: —
Bailey, R. G. The Ecosystem Geogra-
phy of the D eans — Continents. Springer, New York
grs: S. 1995. — in Anatolia, Turkey. Biol.
onse "rvalion 7 72 21:
Bohn, | . 1994. E. ы project for the construction
of a map of the nat e vegetation of Europe at a scale
of 1:2.5 million—lts concept, problems of harmoniza-
tion — ЛУ ation for nature protection, Colloq. Phy-
to 23:
. Ecoregions:
2.
с an mina. 1996. General Map of Natural
Vege tation of Europe. Map (1: 10,000,000). Federal
Agency of Nature Conservation, Bonn, Germany.
Borhidi, A. 1991, Phytogeography and Vegetation Ecology
of Cuba. Akadémiai Kiad6, Budapest, Hungary
Briggs, J. С. 1974, Marine Zoogeography. MeGraw- Hill,
New
ork.
Brinton, E. 1962. ne distribution E Pac — кашан
Bull. — Inst. Oceanogr. 8: 51-27(
Brooks, R. R. 1987. Se rpentine and a Veg >getation: A Mul-
У Approach. Ecology, Phytogeography and
Physiology Series, Vol. 1. Dioscorides Press, Portland,
gon.
Brooks, T. M., S. L. Pimm & N. J. Collar. 1997. Defor-
estation predicts the number of threatened birds in in-
sular S . i : 383-394,
Brown, 1994. Biotic atari South-
western United States and Northwestern Mexico. Univ.
Utah i Salt Lake ep
Bryant, D., E. Rodenberg, ' : D. Nielsen. 1995.
( E. al Risk: 2 m "x ‘af ien Development-
related Threats to Coastal. Ecosystems. Indicator
Brief, World Resource Instit >, о D.C.
. D. Nielsen & L. Tangley. . The Last Frontier
Forests: Ecosystems and Economies on the Edge. World
Resources Institute, Washington, p
— 1 — ү McManus & Spalding. 2000.
Ree ^ at Risk: A Map-based Ei сы of Threats to the
— um s fs. World Resources Institute, Wash-
ing
BSPIC Vi Ne ЛУ RI/WWF. 1995. A Regional Analysis of
Geographic Priorities for Biodiversity Conservation in
atin America and the Caribbean. A Re port for US; AID.
о sea) Program, Washington,
Bullock, Medina (editor).
1996. л, Dry Tropic al Forests. С cambridge Univ.
ress, Cambridge.
. J. D'Amico, is
—
=
М
A
H. A. Mooney & E.
t HOUR & D. M.
: A Con-
servation Se Island Bue Washington , D.C.
C hange ‘hun Institute of Geography and Chinese Ас ademy
of Sciences. 1990, Hi C Atlas of China.
чепсе Press, Beijin
C — Vegetation Map ЕРЕ Committee. 1979.
Veget tation Map of China. Map (1:10,000,000). Science
— го,
;onservation
. 1992 ». Is ай" а — oll sity cline in
the — a Ecol. Evol.
Cody, M. L. Dive he ra ян pe conservation in
Me ана — regions. Pp. 122-152 in M. E.
Soulé (editor), Conservation Bio logy: The Sci ience of
Scarcity and Diversity. Sinauer, Sunderland, Massachu-
setts.
Collar, N. J. & S. N. Stuart. 1988. Key Forests for Threat-
ened Birds in Africa. International Council for Bird
Preservation Monograph No. 3. S-Print, Cambridge,
U.K.
Conservation International. 2000: Tropic ‘al forest. wilder-
tt nservation.org).
эш D.C.
e ntal impacts in
. Mexico: A commentary. J.
9:
ness areas. Web Site:
Conservation Inte emational,
( ORR en s. 1 97
gm,
5
—
pls
); 85—88.
1992. The Mammals of the
( l.
“aly — tud Univ. Press, Oxfor
.M.& (
Cowling, |. Hilton-Taylor. 1994. Patterns of plant
— and — in — Africa: An over-
view. Pp. 31—52 in B. luntley (editor), Botanical
Dive a in Southern is ‘a. NE Botanical Insti-
lute, Pretoria, South. Africa.
Volume 89, Number 2
2002
Olson & Dinerstein 219
The Global 200
Г. Hoffman & C. Hilton-
‹ diversity in southern
Нал (editor), Biotic Di-
Concepts and Conservation.
South Afric:
M. Simmons.
Gibbs Russell.
~
a
Tavlor. s
Africa. Pp. 19-50 in B. J.
versity in Southern Mrica:
Oxford. Univ. Press, Cape Town.
. 1. A. W. MacDonald & 1 1996.
The Cape Peninsula. South Africa: Різне aphical, bi-
ological and historical background to an —
hot-spot of biodiversity. Biodive rsily and Conservation
5: 521-5
Croom, M. . Wolotira & W. Henwood. 1992. Pro-
Subdivisions of the North East Pa-
m. National Ocean Service, Washing-
ada—West Coast
British Colum-
Environment Ca
on.
Coordination Office. North Vancouver.
‚ 1986. Review of the Protected Areas System
in 0c eania. IUCN. Gland. Switzerland. and UK/UNEP,
airobi, Kenya.
а R. E 1974. Biotie Provinces of the World: Fur-
er Development of a System for Defining and Classi-
Hen Natural Regions for shes es of Conservation,
ICN lo. 9. International Union for
Conse rvalion of Nature and Natural Resources. Morge
Occasional Paper No
zerlan
„ Webster,
1995.
estrial Ecoregions
World Bank.
Dine йен, E. D. M. Olson. D. J. Graham. A. |
S. A. Primm, M. Р. O & С.
Conservation
Ledec.
of Latin America a aind the Caribbean. The
Washington, D
Dobson. A. P. J. к Жой; W. M. Roberts & D.
Vilcove. 1997. E distribution of endangere :
species in the United States. Science 275: 550—553.
Ecological Stratification Working Group. 1995. A National
Касе Framework for Canada. Agriculture and
Agri-food — Research Branch, Centre. for
i i urces Research: and Environment
— nt Directorate; Ecozone
Analysis quid наалы Canada.
Land
Elder. D. € J. Pernetta (editors). 1991. Oceans. IUCN.
Mitchell Terba London.
European. Topic Centre on Nature Conservation. 2000.
The Digital Map of European Ecological Regions
(DME ER). Muséum — d'Histoire Naturelle, Par-
is. Franc
Finch. С. (editor). 1996. Mongolia’s Wild Heritage. Mon-
golia Ministry for Nature and the Environment, UNDI
GEF. WWE. Avery Press, Boulder. ата 0.
uinea. The Austra-
Queensland.
Braziller.
1900. Mammals of New
Flannery, T.
ian mo Em Robert Brown.
. 19
New
. The Future Eaters. George
ak:
Forey, P. L.. . Humphries & R. L Vane-Wright (edi-
tors). — [m malies and Conservation. К valuation.
The Systematics Association Special Volume No. 50.
Clare айап Press. —
Freitag. H. 1971. Studies in the ni oan vege tation of Af-
vie кН Pp. 89-106 in m H. is, Р. Harper &
1. C. Hedge (editors). Plant Life : P Asia. The
Botanical Society of Edinburgh. —
Frest. T. J. & E. J. Johannes. 1991. Present and Potential
inse Mi illu iscs Occurring ithin the Range of the
orthern Spotted Owl. Report prepared for Northern
— i Pus rv Team. USES, Portland. Oregon.
Gage. J. D. A. Tyler. 1991. Deep-sea Biology: A
Natural B of Organisms at the Deep-sea Floor,
Univ.
Cambridge Press. Cambridge.
Gallant, A. L.. E. E Binnian, J. M. e & M. B.
Shasby. 1995. Alaska. U.5
Survey Professional Paper — U.S. Government
Printing LR ie m
Gaston, K. J. 4 Williams. Ten Mapping the world's
spe — d taxon approach. Biodiversity Lett
1: 2-8.
Gentry, А. 1993.
Ecoregions of Geological
Diversity d floristic ru of
Neotropical dry forests. Pp. 146—194 in . Bullock.
H. A. Mooney & E. VAN (editors). Se pali Dn
Tropical А Cambridge Univ. Press. Cambridge.
Goulding, M.
The Fishe s and the Forest: Explora-
tions in — Natural History. Univ. California
Press, im kelev
N. J. H. fuilh & D. J. Mahar. 1996. Floods of
Along the Amazon. Co-
Ecology and Economy
P ress; New Yor
Deep-sea b nthic biodiversity. Bio-
Portum:
lumbia Univ.
Grassle, J. F. 199
р.
Maciolek. 1992. Deep-sea species rich-
ness: — and local diversity estimates from quan-
Amer. Naturalist 139: 313-
Science
titative bottom
341.
Gray, J. S.
conservation needs.
samples.
Patterns, threats anc
and Conservation 6:
‚ Marine biodiversity:
Biodive тэу
53-175.
Groombridge, В.
& M.
Seas:
Jenkins (editors). 1996. =
Diversity of the Я Regional Approach. WEN
Biodive silty Series No. 4. World Conservation Ment
ing Ce nire, — К gu c Press; Cambridge. .
. Freshwater Biodiversity: A e re-
mir Glob: ib Assessment. WCMC Biodiversity Series
8. World Conservation Monitoring Centre. Cambridge.
ЕРИНИ ‚С. L. Valutis; D. Vosiek, В. Neely. К. Wheaton.
Je Тош: àl & B. Dich Is. 2000. Designing a Geography
of Hope: A — for
Conservation Planning. Special ed.. 2 vols. The
— Virginia.
M. Adams & W. H. Martin (editors).
992. B А of the Southeastern United States:
* uatic Communities. John Wiley and Sons; New York.
Hamition, А. C. & R. Bensted-Smith. 1989. Fores
i Fasl Usambara Mountains. Tanzania.
Practitioners Ecoregional
he Nature
Conservancy,
Ыт Кпеу.
Hamilton. L. S.. d. 0. atena (editors).
1995. — a) Montane С loud Fo oresls. d Teologi al Stud-
ies 110. Springer-Ve rlag. New York.
. Billington (editors). 1990.
Г Tropical и The Amer-
IUCN, WCMC, Cam-
Harcourt. €. S.. J. —
The Conservation
World буена Union.
—
idge
Hayden. B. . Ray & R. Dolan. 1984. Classification
of coastal i marine environments. Environm. Conser-
— 11: 199—207.
. B. 1996. Historical mogeography o of * In-
dian verte brates. e Rev. Ecol. Syst. 27 ‹
Hernandez. Н. R. T. Вагсепаз. 1995.
cacti in the € ae ihuan De sert: Ll Distribution patterns.
1
Conservation Biol. 9: 1176-1188.
. 9
Hedges,
Hilbig. W. 1995. The Vegetation of Mongolia. SPB Aca-
demic Press. Amsterdam. The Netherlands
Hocutt, С. H. & E. ( ¡ley (editors). 1986. The Zooge-
ography of orth — ап Freshwater n ishes. John
Huntley. B . J. (editor) "1999. Biste Diversity in Southern
220
Annals of the
Missouri Botanical Garden
Africa: Concepts and Conservation. Oxford Univ. Press,
Cape Town, South Africa.
— — (e ditor). 1994. Botanical Diversity in Southern Af-
rica. National Botanical Institute, Pretoria, South Afri-
ICBP 1992. Putting Biodiversity on the Map: Priority Ar-
eas for Global Conservation. Inte rnational Council for
Bird Preservation, Cambridge, U.
IUCN. 1988. Coral Ree da of ihe "World.
WCMC, € Sunbridge. |
. 1990. Biodive аа in Sub-Saharan Africa and its
СБа. Occasional Papers of the IUCN S
vival Commission No. 6. IUCN
Gland, Switzerland.
‚ 1991. The Conservation Atlas of Tropical Forests:
Asia sud the Pacific. MacMillan, London.
— ——. 1992. The Conservation Atlas of Tropical Forests:
ux Millan, London.
IUC NIUNEP_ 1986a. Review of the Protected Areas Sys-
tem in the Afrotropical Realm. IUCN, Gland, Switzer-
and.
3 vols. IUCN
and
Sur-
Unit,
)ecies
Publications
— 30b. Review of the P — Areas System in
the — Malayer Realm. IU( Gland, Switzerland.
IUCN/UNEP/W WF. 1987. — ar: An Environmental
Profile. IUC М, G — Switzerland.
Johannes, R. E. & B. G. Hatcher. 1986. vid tropical
marine environme * Pp. 371—392 in М. yulé (ed-
itor), Conservation Biology: The Science of m arcity and
Diversity. Sinauer, Sunderland, Massachusetts.
Johnson, T. H. & A. J. Stattersfie = 1990. A global review
of island endemic birds. Ibis : 167-180.
Kalin rm M. T., P. H. 4 эж = M. D. Fox (editors).
1995. Ec ology and Biogeography of Mediterranean Eco-
Sus ms in Chile, California, and Australia. Springer-
erlag, js w York.
— J. T. & C. Meacham. 1999, Synthesis of the
North ral rican Flora, Database. Version 1.0. North
— Botanical e C — Hill.
Kelleher, G., C. Bleakley Wells. 1995.
Represe — System a — Protected |.
vols. Great Barrier Marine Park Authority,
hi —
Fa World
Bank, IUCN, Washington, D.C
Kellogg, E., J. Weigand, A. Mitch ell & D. Morgan. 1992.
Coastal Temperate Rain Forests: Ecological Character-
istics, маи and Distribution Worldwide. Occasional
Paper Series No. ;cotrust
— Put Orekan:
Kendall, & M. Aschan. 1993. Latitudinal gradients
in the struc — of macrobenthic communities: A com-
— of Arctic, temperate and tropical sites. J. Exp.
r. . Ecol. 172: 157-169.
Kingdon, T 1989. Island Africa: The Evolution of Africa’s
Animals and
Prine ier
997. African Mammals.
and Conservation Inter-
Plants. Princeton Univ. Press,
Academic Press, San
Disgo
gy A. 1987. Pun Natural History of the USSR.
w-Hill, New
Rotts on м. & T. mos п. 1996. Asia-wide Assessment
of Freshwater Biodiversity. World Bank Technical Paper
Number 281. The World Bank, e d C.
Krever, V, E. Dinerstein, Olson & L. Williams.
1994. Conserving Russia's — a Dive rsity: An An-
лш al Framework and Initial Portfolio.
е D.
Kuma S. F. 1990. — regionalization of the USSR.
Map (1: DA. Sheet 15 of Forests of the USSR.
M
Investment
Main Division for — and Cartography of the So-
viet of Ministers,
Lacerda, L. D. (с a a 1993. Conservation and Sus-
amable Utilization of Mangrove Forests in Latin Amer-
ica and Africa Regions. Parts | and 2.
Society for Mangrove Ecosystems and
Tropical Timber са Okinawa.
ee, D. S., C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D.
E. McAllister & J. R. a Jr. 1980. Atlas of North
American Freshwater Fishes. North Carolina Biological
Survey
L —
—
—
International
International
‚ 1997, jug беш Dynamics and Conserva-
> Freshwater Fish of Africa. Cambridge Univ.
Press, pon ae tta LK.
—— D. Paugy & G. G. Teugels (editors). 1992. Faune
des Passani d'Eaux Douces et Saumátres de l'Afrique
IC
tion:
de Ouest. Tome 2. — TOM
Lieske, E. & R. — р Reef Fishes. Prince-
ton Univ. Press, Pr rinc
Longhurst, A. 1998. Ec — al Geography of the Sea. Ac-
ademic P ress, London.
Lovett, J. C. & S. К. Wasser (editors). 1993. Biogeography
and E — of the Rain Forests of E astern Africa. Cam-
bridge Univ. Press, Cambridge,
Lowe-McConnell, R. H. 1987. Ec logic ‘al Studies in Trop-
ical F a Саш Cambridge Univ. Press, Cam-
bridge, l
Mac B. T & N
ress, C ambridge
. Hicks.
„ Massa ‘tts.
— ——— & G. Bunting. 1996. ууш ТЭЙ Natural Habitats
of the Indomalayan — ıl Database. World
Conse p „Moniióring Centre, Cambridge,
MacNae, W. 1968. A general account of i — and
flora of mangrove swamps and forests of the Indo-West-
Pacific re — Adv. Mar. Biol. 6: 73—270.
Maggs, С. . H. Kolberg & C. J. H. Hines. 1994.
НИ dise ity in Namibia. Pp. 93-104 in B. J.
Huntley (editor), Botanical Diversity in Southern Africa.
National Botanical Institute, Pretoria
artin, С. 1991. The Rainforests of West Afric y Ecolo-
19 Duca C ees —
Martin, W. H., . Boyce & A. Echternac bh (editors).
1993. — of Mie Sout us astern United States:
Lowland Terrestrial Communities. John Wiley and Sons,
б York
Maxwell, ia C.J, — M. E. Jensen, S. J. Pau
tian. H. ae & D. M. Hill. 1995. A Hie —
ا of Aquatic ee ‘al Units in North Amer-
1 (Nearctic ты, USDA Forest Service General
Тес мк — NC- 176. St. — —
McAllister, ‚ F. W. Schueler, C. M. Roberts & J. P.
Hawkins. А Mapping and G É analysis of the global
distribution of coral reef fishes on an equal-area grid.
Pp. 1 К.Т. “Miller (editor), Mapping the Di-
‘sity of vv Chapman and Hall, London.
McClanahan. T. R. & T. P. Young (editors). 1996. East
African E and Their Conservation. Oxford
Univ. Press, New Yor
McDowall, R. (editor). 1996. Freshwater Fishes of South-
eastern Antal Reed Books, Chatswood, Australia.
Mitchell, Fragile Paradise: Nature and Man
in the ке C ollins, London.
Mittermeier, R. ` B. Werner & 1996. New
Caledonia—A orion imperative for an ancient
land. Oryx 30: 104-112.
>. Mittermeier & P. R. Gill. 1999,
Earth’ s Biologically Richest and Most Endan-
1996. Wild China. The MIT
>
Basel.
А. Lees.
. N.
Hotspots:
Volume 89, Number 2
2002
Olson & Dinerstein
The Global 200
gered Terrestrial Ecoregions. CEMEX and Conservation
— Washington JC.
— A. 1975. Outline of Japanese vegetation. Pp.
1 K. Nana К. Yoshida & М.К
a as in Conservation of Natural Terrestrial Ecosys-
" ms in ZA JIBP Synthesis Vol. 8. Univ. Tokyo Press,
. Jap
— Ministry for Nature and the Environment, Unit
* ations De -velopme nt Programme (UNDP)/G "m
Environme ent dy ility (GEF). and World Wide Fund for
1996. Mongolia’s Wild Heritage. Avery
S. Poe Colorado.
Mori, T. 1936. Studies on the Geographical Distribution
of Freshwater. Fishes
Univers sity, v Vor n. poe
Myers, G.S.
and the e Do T higher taxonomic
lution 14: 323-33:
ato (editors).
zu
| Eastern. Asia. Keijo Imperial
. The jene mic fish fauna of Lake Lanao.
categories. Evo-
— чег, C. G. Mittermeier, G. A. B.
Kent. 2000. Biodiversity hotspots for
858
vers, N., Ag A.
da Fonseca & J.
conservation priorities. Nature 408: 853—
Nagelker Н . M. V. Mina, T. Wudneh. F. A. Sib-
bing & qo Dass. 1995. In Lake Tana, a unique
fish fauna — protection. BioScience 45: 772-775
New Zealand Department of Conservation. 1987. Ecolog-
ical Regions and Districts of New Zealand. Map (1
500.000). Department of Conservation, Wellington, New
Zealand.
Nichols, J. Л. 1943. The Fresh-water
Americ due i Muss um of Natural History,
l.
(12—44 id.
Fishes of China.
Ne "Ww You К
Nicol & О. — 1989, Madagascar: Revue
de la Es onservation et des Aires Protégées. WWE Inter-
National, : — Switzerlan d.
Noirfalise ap of the Natural Vegetation of the
Me — ountries of the European Community and the
Council of Europe. Map (1:3,000,000). Council of Eu-
rope, жае France.
Norse, К. A. 199 ‚ Маййайв the world’s marine biolog-
ical 7 a Bull Mar. Sci. 57: 10—13.
Oates, К. Apis Primates: Status Survey. and
Conservation p tion Plan, revised ed. IUCN, Gland,
Switzet ia
Oberdorff, T.. J.-F. Guegan € B. Hugueny. 1995. Global
scale us rns of fish species richness in rivers. Ecog-
raphy 18: 345-352.
Olson, D. M. & E. Dinerstein. 1998. The Global 200: A
represe ntation approach lo conse туш the Кап» vu
biologically valuable ecoregions. Conservation Biol.
502—515.
A Con-
America
. Cintron & P. lolster. 1996.
servation — sssment of Mangroves of Latin
and the Caribbean. Conservation Science
WWF-US. oe D.C.
Program,
— ———, В. Chernoff, C. Burgess, I. Davidson, P. Canevari,
Dinerstein, "Gt Castre Morisset. R. Abell & E.
e 1999, Fre ut T dodici of Latin America
and the Caribbean: A Conservation Assessment. WWF-
US, Wetlands International, Biodiversity Support. Pro-
gram, е USAID, W: — D.C.
——— F. Dinerstein, E. ici E И . D. Bur-
gess, e V. N. Y vell. E. C. U nde rwo * a . D'Amico,
* Morrison, € . J. Louc T. F. АП.
| H. Rie kets, I. Tue e W. Wetten-
Pp. Hedan & K. Kassem. 2001.
мамы мину ч ths world: A new map of life on Earth.
BioScience ql :
Omernik, J. M
Terrestrial
933-938.
1995. bud Ш Ecoregions of the Conti-
National Academy, Washington,
— — 199
nental US. National — and Environment Effects
Research
aborato S Environmental Protection
Agency, Washington, n. C.
Oosterbroek, 1994. Biodiversity of ^x Mediterranean
region. Pp. ege in P. L. Forey, C. J. Humphries &
{ Vane-Wright (editors), Syste matics and Conserva-
tion EvalusBon.
Vol. No. 50. Clarendon Press, —
Orians, G pim Endangered.
Applic 200-20
башы. E
The Systematics * iation Spec ial
~
den level? Ecol.
J Gage & M. V. Angel —
1997. Marine p е rsity: Patterns and Proc
и Univ. Press, Cambridge, U.K
Parker, ine iu J. L. Carr (editors). 1992. Status of
— Ren — in the Cordillera de |
Southwestern Ecuador. RAP Working
Papers 2. Conservation Inte rational,
jacent Areas of '
Washington, D.C.
. H. Ge * R. 'oster, L. H. Emmons & J.
V. M msen, Jr. 1993. The land Dry Forests of Santa
Cruz, Bolivia: A d Tobal Conservation Priority. RAP
Working Papers 4. Conservation International, Wash-
ington, D.C.
А ОР. Jr. 1991. The I
Timber Press, Portland,
Peterson, ^. T. О. es- Villela,
Llorente- hocque Is, M. A.
evs rro-Sigiienza,
Pines of Mexico and Central
id us
LS. — Paniagua,
p Martinez, A. G.
Torres-Chávez
—
| & l. —
Fernández. 1993. Conservation. priorities in. Mexico:
— ap in the world. Biodiversity Lett 33-38.
Picker, M. D. & M. J. Samways. 1995. Final diversity
and endemicity of the Cape Peninsula, South Africa: A
irst assessment. Biodiversity and Conservation 5: 591—
6006.
Pierrot-Bults, A. С. 1997,
Biological diversity in oceanic
mac rezoaplankton More than counting species А
93 in R. & М.
^p.
. Ormond, J. D. Gage ¢ V. Angel,
Marine duced sity: — ‘rns s and Processes. Cambridge
iniv. Press, Cambridge,
І
gral & D. A. Keith. 1994. New
procedures for reserve selection in New South Wales:
Pressey, R. L.,
Maximizing the chances of achieving a representative
network. Pp. 351-373 in P. L. Forey,
< R. I. Vane-Wright (editors), Systematics and Conser-
vation Evi — The
J. Humphries
> Systematics For lation Speck i
ol. No. 50. Clarendon Press, Oxford.
о Mos K. 1991.
| File, Oxford, U.
был T. P, К. Bye,
Madagascar: A — History.
—
^. Lot & J. Fa (editors). 1993.
Biological Diversity of Mex
tion. Oxford Univ.
Raven, P. H. 1988. Our diminishing topi cal forests. Pp.
п 9-122 in E. O. Wilson (editor), Biodiversity. National
ademy, eec d
nite С. (
oceans. » 36—
ico: Origins and Distribu-
Press, New Yor
~
88. Ecologic ‘al dive ersity in coastal zones and
50 in E. О. Wilson (editor), Biodiversity.
C.
Coastal-zone biodiversity patterns. Bio-
— 11: 490—498.
& B. P. Hayden. 1993. Marine biogeographic
коө of the Bering, С -— ‘hi, and Beaufort Seas.
Pp. 175-184 in K. Sherman, L. M. Alexander & B
Gold байо, Large Marine — Patterns, Pro-
Yields.
vancement of W ience, Washington, D.C.
Reaka-Kudla, M. L., D. E. Wilson € E. O. Wi
tors). 1997. M Rs Il: Understanding and Protect-
—
cesses and American Association for the Ad-
Ison (edi-
222
Annals of the
Missouri Botanical Garden
ing our Biological Resources. Joseph Henry. Press,
Washington, D.C.
Reeves, R. R. & — Dolphins, Por-
Action Plan for the
Gland, S See awn
5. —€—
poises, and Whales. 1994—19
Conse dos of Cetaceans. IUCN,
Ric ke tts, . Dinerstein, D. > Olson, C. J. Lou
. Eic РА p. x llaSala, K. Kavanagh, P. — Р
T. Hurlev, Carney, R. Abell € S. Walters. 1999
‘ial Ecoregions of North ; : А Сопзегуа-
1 Press, Washington, D.C.
Terrestr
tion Assessment. Islan
E. & ROE
mangrove floras. Рр. 215-
Ricklefs & D. Schluter (editors), Species pote in
Ecological Communities. Univ. Chicago Press, Chicag
Robbins. C. R. 1991. Regional dive ape among Caribbean
fish species. Bios ence 47: 458—
Roberts, T. R. 1975. Е val — vin 'an
freshwater fishes. Zool. Ann. Soc. 57: 249-:
Sherman, K., 1 Gold —
1990. Large Manne Ecosystems: Patterns,
and Yields. American Association for the Advancement
of Science, е С.
Silva, J. М. С. 1998. Um Método рага о ipei 'imento
de Áreas Prioritárias para a Conservação na Amazonia
um Report pre pared for World Wildlife Fund. Bra-
P rocesses
Spalding, M., F. Blasco & C.
Mangrove Atlas.
Ecosystems, Okin
Field (editors). 1997. World
dbi International Society for Mangrove
wa, Japan
State of the — nt Advisory Council. 1996. Austra-
: State of the Environment. De ‘partment of the Envi-
ronment, Sport and Territories. CSIRO Publishing, Col-
Australia.
n
Stattersfield, A. J., M. J. Cros sby, A. J. Long & ae 1. Wege.
1998. Endemic s Bird Areas of the Wo — irdlife Con-
servation Series No. 7. Birdlife но riae zam-
bridge, U.
Stewart, J. M. 1992. The Nature of Russia. Boxtree, Lon-
don.
Stewart-Cox, В. 1995. Wild Thailand. MIT Press, Cam-
bridge, Massachuse tts
— J. A.. P. Bens & B. К. Holst —
)5. Flora of the Ve 'nezuelan Guayana. Vol. 1: Intro-
rii Missouri Botanical Garden, St. Louis
Stiassny, M. L. . K. Sehliwen € W. J. — 1992.
new species Hoe k of cichlid fishes from Lake Bermin.
Cameroon, with a description of eight new species of
Tilapia (1 ^ bruidei: Cichlidae). Icthyol. Explor. Fresh-
waters 3: 311—346
Suchanek, T. H. 16 9 I. Temperate coastal marine com-
munities: Biodiversity and threats. Amer. Zool. 34:
100-114.
Sujatnika, J. P., T. R. Soehartono, M. J. Crosby -
diastuti.
1995. p onserving Indone — Bio liv A. `
Endemic Bird Area Approach. BirdLif
Indonesia : "rogramme, Bogor,
Sullivan, K. . Bust
vation ns 'ssment of the Freshwater Marine Ecore-
nd
‘aribbe ап. BSP/WM
ife we ed
sa sia.
tamante — 1996. A Conser-
gions of Latin America папі the (
TNC, Washington,
Taki, Y. 1975. primary
freshwater fishes in кш — ipal areas of Southeast
0—214
м
—
— distribution. of
Asia. $ Asian Stud. (Kvoto Univ.) 13: 20
Terborgh, 1. & В. Winter. ion \ me Mar for siting patio
and reserves with special reference to Colombia and
Ecuador. Biol. Conservation 27: 45-58.
Thackway, R. & I. D. Cresswell (editors). 1995. An Inter-
Latham. 1993. G — IE rns of
R. E.
Ec alos of Java and Bali.
im р Regionalisation for Australia:
Framework for Setting Priorities in the National Re-
serves System на Program. Version 4.0. Aus-
tralian Nature Conservation Agency, Canberra.
Thieme. M. L., R. A. Abell, D. M. Olson, E. Dinerstein,
M. L. J. Stiassny, J. D'Amico & E. M. Underwood. In
press. Freshwater regions of Africa: A Conservation
Assessment. pes Press, Washington, D.C.
Timberlake, J. R „ Müller. 1994. Identifying and de-
scribing areas ior vegetation conservation in Zimbabwe.
Pp. 125-140 in B. J. Huntley (editor), Botanical Diver-
sity in S Africa. National Botanical Institute,
Pretoria, wy Africa.
Udvardy, M | 1975. World ре Provinces.
CN B al Paper No. 18. Gland, Switzerland.
lisse Academy of Sciences. 1988. Red Book—Russian
Soviet Federal Socialist Re public. 2 vols. USSR Acad-
у of Science, is cow
Vane- Wright R. L, C. J. a & P. H. Wil
1991. What to potect?—Syst ematics and b agony of
qu Conse — 55: 235-254.
быш, 1. N. 1995. Corals in Space and Time: The
goin and Evolution of the Scleractinia. Com-
5 те — — New Yor
. Ice-Geography ‘of Greenland. Map (1:
5. 000,000). Genetisk Institut, Copenhagen, De nmark.
Vinogradova, N. G. 1
liams.
77
LI
me
o
7. Zooge dw of the abyssal and
87 in ` . Blaxter, A. J.
A. V. Gebruk & P. A r. r (editors), Ad-
vances in manne parma bur 32: ‘The Biogeography of
San Diego.
A Funk (editors): 1995. Hawaiian
— on a He и Spot Archipelago
Wagne
Bioge gr —
Earth's
Living Resources. Chapman and Hall, London.
C. & A. J. Lc 5. Kev Areas for Threatened
— in the Neoiropitis Birdlife gc Series
No. 5. BirdLife International, Cambridge.
— М. J. A. (editor). 1978. Hinge pe Ecology
of Southern Africa. Junk, The Hague.
Wetlands International and The I —
lands of International Importance. i
World Bank, Washington, D.C.
i | 33. The Vegetation of Africa: A Descriptive
Memoir to Accompany 1 the UNES SCO/AE! l'FAT/UNSO
Afric a. Natural Resources Research
1996. Wet-
Asia. Map. The
Whitmore, T. C. Hi n Rainforests of the Far
East. Oxford Univ. s ss, Oxford, U.K.
1990. An ntti to Tropical Rainforests.
Clare ndon tress, Oxford, U.K.
& ( ance (editors). 1987. Bioge — —
umen tina in Tropical America. Oxford Mor
ogr. de gee No. : و Press, Oxford, UK.
Whiten: A. J.. M. Mustafa & G. 1987a.
The Ecology : те AR U niv. Press, Jakarta,
Indonesia.
S. J. Damanik, J.
. Henderson.
Anwar & N. Hisyam. 1987b.
The Е — of Sumatra. Gadjah Mada Univ. Press, Yo-
gvakarta, Indo mesia.
R. E. Soeriaatmadja & S. Afiff. 1996. The
The Ec MS of — Se-
ries, — ne Il. Periplus E — s, Singapo
Wiker . (compiler). 19 Te earal Fco 'oregions of
ina и al Land С Pa ‘ation Series, No. 19
Environment Салай, Hull, Quebec.
Volume 89, Number 2
Olson & Dinerstein 223
The Global 200
Wikramanayake, E.. К. — C. Loucks, Olson,
J. Morrison, J. — M. Knighi & Е —
2001. Terrestrial Ecoregions of j Indo- Pac ifie: A Con-
servation
enn,
C
Williams.
киюе» ге aede sss, and endemism in conservation.
E 269-287 in Р. L. J. Humphries
Vane-Wright
E aul tiation
50. Clarendon Press, Oxford.
& R
'orey, С.
(editors). Systematics and Conservation
The Systematics Association ай ial Vol-
ume No.
| ane-Wright. 1991. Measuring
Бейге: € re d 388 for conservation
лое Is
5—07
out Biodiversity — Acad-
2. The Diversity a Life. Belknap Press, Cam-
bridge, е husetts.
WWF/IUCN. 1994—1997.
Guide and Strategy for their Conservation.
LK.
Centres of Plant Diversity: A
3 Vols. IU
Publications Unit, Cambridge, Û
Yim. Y.-J. 1977. Distribution of forest vegetation and cli-
mate in the Korean Peninsula: IV Zonal distribution of
orest vegetation in relation to thermal climate. Jap. J
à — ү
27: “Ж” Li
87 A Zoogeography of some primary
Asia. Walla-
CO
Zakaria- Ismail.
freshwater EN s in South and Southeast
ceana 17: 6-9.
. 1994. Zoogeography and biodiversity of the fresh-
Wate r fishes of Southeast Asia. Hydrobiologia 285: 11
— V. J. & S. V. Gerd. 1961. Fauna and Flora of the
ivers. Lakes. and Reservoirs of the USSR. Smithson-
ian Institution and National Science Foundation, Wash-
ДН
inglor
Zhao. he ng Guangmei, Wang Huadong & Xu Jialin
irs 1990. The Natural History of China. McGraw-
Hill. New York.
Zoharv. M. 197 » Geobotanic al Foundations of the Middle
East. x Is. ustav Fischer Verlag. Stuttgart. Ger-
many.
APPENDIX l. WEIGHTING AND MEASURING BIOLOGICAL
DISTINCTIVENESS CRITERIA
The we ighting and measurement of the parameters used
1 s of terrestrial ecore-
to assess iological distinctis
gions of Nort i merica are d here to detta
how different biodiversity features were evaluated as con-
servation targets and how analyses were tailored to differ-
ent biomes. Comparisons among biodiversity. parameters
were only conducted within the set of ecoregions sharing
the same biome.
SPECIES RICHNESS ®
G lobally outstanding 100
ligh lo
Medium 10
A
*Only nalive species were used in species counts.
ENDEMISM
Globally outstanding 100
High 25
Medium 15
W 5
For species richness and endemism, the total number
he
of species that occurs within each ecoregion, and the total
al ‘Species —
and native vascular plants, regional experts
— Балу Roth analyzed land snail Histon for
western and ed rich-
i John
ness — endemism —— for native v
ort th A meric a,
cies was consider ndemic^ to an ecoregion if its es-
imated range fell « entire сн within a single ecoregion (strict
vithin a single
endemic), 75% or more of its range fell y
ecoregion (near-endemic), or its range was less than
50,000 km” (range-restricted). If a species had a signifi-
cant distribution outside of the United States and — а,
il was nol conside red as an ende mic. Higher taxonomic
uniqueness—e.g.
cies or communities, primitive lineages—was :
ered for identifving globally outstanding ecoregions from
an endemism perspective.
The actual number of species and endemies for each
laxon found within an ecoregion were log transformed to
uce the influence of very species-rich groups. The logs
> then summed to derive a single richness and ende-
‘hese scores were plotte d for the ec oreglons
<
mism score.
within each biome and the curves broken subjectively into
high, medium, and low scores. Globally outstanding scores
were determined through
ecoregions within the same biome found throughout the
comparisons with values for
work
UNUSUAL ECOLOGICAL OR EVOLUTIONARY PHENOMENA
Globally outstanding 100
Regionally outstanding 2
No globally or regionally unusual phenomena 0
e P, nm J
amples of unusual оне ев volutionary phe-
|
Ex
nomena al global or гери за SC ‘ales i me lie re alive M in-
tact,
caribou. intact predator a
centrations of breeding waterfowl and shorebirds. extraor-
arge-scale migrations of large vertebrates such
em ibl: age 8: supe abundant СОП"
dinary levels of adaptive radiations, rain-fed flooded grass-
lands on limestone, and conifer forests dominated. by
gigantic trees.
GLOBAL RARITY OF BIOME
Global rarity 100
Regional rarity 5
Not rare al global scale 0
ts that were considered globally rare
woodlands, an
Biomes or habitat
editerranean-climate forests.
include
тар. te mperale rainforests. and paramo.
TOTAL SCORES FOR DETERMINING BIOLOGICAL
DISTINCTIVENESS INDEX
The points from each criterion were summed to arrive
at a final score. This score was then translated into a bi-
ological distinctiveness category as follows
Globally outstanding 15, А or 55+ points
Regionally outstanding 30, 35. 40
Bioregionally outstanding — 20. T
Locally important
224
Annals of the
Missouri Botanical Garden
Ecoregions identified as globally outstanding were sub-
sequently compared with similar ecoregions around the
world to validate their hit status.
APPENDIX 2. ASSESSING CONSERVATION STATUS OF
ECOREGIONS
Conservation status measures landscape and ecosys-
tem-level features and relates inet to the ecological in-
tegrity of with increasing habitat
loss, degradation, and bun ia ecological processes
cease to function naturally, or at all, resiliency to distur-
bance declines, and major components of biodiversity are
steadily eroded. We a conservalion status «
ecoregions in the tradition of IUCN Red Data Book cate-
gories for threatened and endangered species: critical, en-
dangered, and vulne га For ес — we — Hie fol-
lowing | conservatior мац categories: ritical,
endangered, vulne e re dis sly zie and re Чай ly
intact. Throughout all of the regional analyses, the specific
parameters and thresholds used for assessing conservation
status were tailored to the characteristic patterns of a
у, Kin ‘al dynamics, and responses to distu
bance of different biomes.
ecoregions, namely, how
assess the
diverait
TERRESTRIAL ECOREGIONS
We present the method used to assess conservation sta-
tus for the terrestrial ecoregions of North America to il-
lustrate the approach (Ricketts et al., . The re ае
contributions of different
4096—habi ze of remaining
blocks of — o 20%—de gree of habita fragmen-
tation, and 1
servation status
apshot con-
vel of en-
dangerment. The point thresholds for different categories
of conservation status were as follows: critical 89-100
5—88, vulne rable 37—64,
РА
ج
=
b
|
.
were determined by summing points — for each ра-
rameter. Individual parameter point values were associ-
©
ated with different landscape scenarios. For example, —
habitat loss scenarios were related to points as follow
% Original habitat Heavily altered Ме
90-100% 40 20
75-89% 30 15
50-74% 20 10
10-49% 10 5
9% 0 0
3
3
w
A
©
E:
S
es
£z
ص
—
=
— habitat (10 pointa), and therefore 10% intact hab-
. By combining the two scores, the ecoregion would
receive a total score of 20 polola: "рите rent quantitative
and qualitative biodiversity and landscape ecology char-
acteristics are used to define intact, altered, and heavily
altered states tailored to the specific patterns and dynam-
ics of different biomes. Total scores for each of the param-
eters are summed to give a total conservation status index
SCC ОР:
=
apshot scores were subsequently modified by a 20-
s assessment
as facing high threat were elevated to a
servation status. The threat analysis estimated the cumu-
lative impacts of all current and projec = threats on hab-
habitat degradation, and wildlife
exploitation using a point system assoc — with different
qualitative and quantitative impacts. Using an n — 0—
100 points, pending threats within an ec oregion
sessed and point totals assigned for each of f the > ae
cate gories. c onversion threats were considered to be the
most serious, and thus habitat loss c omprised half (50) of
all possible points in the weighting of threats. For exam-
ple, 50 points were assigne
or more of remaining habitat would be categorize
heavily altered within 20 years. For conversion of be sit
10% and 24% of remaining enn a score of 20 points
was assigned. The remaining two threats, е ——
dation and wildlife e — ation, were assessed u
imum point totals of 30 and 20 respectively ен asc *
based on high, а edi or no threat.
itat conversion,
>
EL GENERO JUSTICIA Cecilia Ezcurra?
(ACANTHACEAE) EN
SUDAMÉRICA AUSTRAL!
RESUMEN
El género Justicia es el más grande y — de la familia Acanthaceae, y está muy diversificado en las regiones
iste trabajo с ompre nde la revisión de las espec ies de Sudamérica austral
Las 38 especies de Justicia de esta región se agrupan en 8 secciones
cápsula y semilla. Todas las especies
e da una clave para
tropic Шш y subtropicales de América del S
sobre la base de material de Argentina у — ay.
en relación a caracteres de su morfología. espec ialmente de inflorescencia, flor,
se describen, se dan referencias de ilustraciones ya publicadas o se ilustran por primera vez, y
s para la region, y la nomene — de las mismas incluye varias sinonimias
identificarlas. Algunas no habían sido citadas
rales. También se dan notas sobre su fe nologí . hábitat, distri-
nuevas basadas en el estudio de los tipos nomenclatu
bución geográfica, caracteres distintivos y afinidades taxonómicas.
ABSTRACT
The genus Justicia is the largest and most complex of the Ac — ‘eae family, and is highly diversified in tropical
and bolo ‘al regions of South America. This work revises the species of southern South America based on material
from Argentina and Paraguay. The 38 spes ies of Justicia that are found in this region are grouped in 8 sections in
relation to their — ‘al characters especially inflorescence, flower, capsule, ed seed morphology. All the spe cies
are described, references to published illustrations are given or they are illustrated for the first time. and i is
provided for their identification. Some had not bee n reported pre ‘viously f for the region, and their nomenc Шз, inc —
several new synonymies based on the study of type material. Notes on phenology, habitat, geographic distribution,
distinguishing c characters, and taxonomic affinities are also give
Key words: Acanthaceae, Justicia, morphology, taxonomy, southe rn South America.
El género Justicia L. es el más grande y com- (Burkart, 1943) u ornamentales (Bailey, 1949; Pa-
rodi & Dimitri. 1980). Muchas de sus especies tam-
plejo de las Acantáceas a nivel mundial, y el de
bién tienen importancia ecológica por ser elemen-
mayor número de especies en los regiones tropi-
cales y subtropicales del continente americano. tos abundantes en el sotobosque de selvas у
Este género comprende aproximadamente 600 es- bosques húmedos, o por ser frecuentes o dominan-
y arbustos perennifolios (Graham, tes en ambientes semiáridos. En México está
pecies de hierbas
1988). varias de ellas importantes como forrajeras presentado por alrededor de 75 especies (Daniel.
! Agradezco a Tom — | por е srme cedido generosamente el estudio de las Acantáceas de Paraguay que origi-
defines nte. pensaba realizar él. Mi reconocimiento a Angel Cabrera por sugerirme este tema hace varios años,
Vladimiro Dudas por las ua iones, a Ana Ladio y — Amos por
ejemplares e índices, y a los curadores y personal de los siguie ntes he rbarios por facilitar material para este estudio:
BM, CORD, CTES, E, ЕСО. G, K, LIL, LP. М, P NY. MO, SI, US, W. También agr adezco a Fátima Mereles, HM
Soria su valiosa ayuda en Parto tanto en el herbario (FCQ) como en el campo.
colección y sus interesantes datos sobre distribuci ion
a asistencia en la edición de las listas de
Basualdo. Rosa Diesen у Nélida S
a Carlos Saravia Toledo su invalorable colaboración en viajes de
y ecología de las Acantáceas del Chac
"te trabajo se realizó en parte con una beca «
Tinker para realizar — en el Jardín Botánico de Missouri, St. Louis,
realizar estudios en el National Museum of Natural History. Washington, y una beca del Jardín РН о de Kew (cor
financiación de Mellon Foundation) para trabajar en herbarios europeos. También el proyecto PIP CONICET 1 20/96.
financió parte de mis visitas a herbarios de Argentina. Agradezco a todos los que me ayudaron en estas instituciones,
Dick Brummitt, Enrique Forero, Kaj Vollesen y Dieter Wasshausen. Mi reconoci-
Missouri Botanical Garden (con financiación de la Fundación
una beca de la Smithsonian Institution à 1
y muy especialmente a Luis Ariza,
miento también a los que de m y otra manera me estimularon para que ea, este trabajo. en especial a Vicloria
Graham y Lucinda Me Dade. Y mi especial consideración a los tres revisores anónimos y a los editores que gracias a
su detallada y rigurosa larea me — a — Mi trabajo es financiado por el Consej jo Nacional de Investi-
gaciones Científicas y Técnicas (C ONICE T) de Argentina a través de la Carrera del Investigador a la que pertenezco.
? Departamento de Botánica, Centro Ee ded Universitario Bariloche, Universidad Nacional del Comahue, Quintral
1250, 8400 Bariloche, Argentina
ANN. Missourt Bor. GARD. 89: 225-280. 2002
226
Annals of the
Missouri Botanical Garden
1993), en Colombia por 85 (Leonard, 1951—1958),
en Ecuador por 27 (Jörgensen € León-Yánez,
1999), en Perá por 50 (Brako & Zarucchi, 1993),
y en Argentina por 28 (Ezcurra, 1993a, 1999).
Hasta ahora no existen trabajos sobre la siste-
mática de este género en la porción austral de
América del Sur, excepto algunos tratamientos re-
gionales (Lindau, 1894, para Argentina; Rambo,
1964. para Río Grande do Sul. Brasil: Dawson,
1965 y 1979, para Buenos Aires у Entre Ríos, Ar-
1969, para Santa
Catarina, Brasil; Ariza-Espinar, 1971, para el cen-
gentina; Wasshausen & Smith,
tro de Argentina; Ezcurra, 1993a, para Jujuy. Ar-
gentina). Debido a esto se propuso realizar un es-
tudio de las especies de Argentina y Paraguay,
países donde el género está especialmente repre-
sentado y presenta una diversidad importante.
El objetivo principal de este trabajo es describir
la morfología y resolver la nomenclatura de las es-
pecies de Sudamérica austral, lo que facilitará la
identificación de la mayoría de las especies de la
región subtropical y templada de América del Sur.
Además este estudio comprende un análisis mor-
fológico teniendo en cuenta especialmente la ar-
quitectura de la inflorescencia, la morfología de la
corola, y la forma y superficie de la semilla, carac-
teres que ya han sido sefialados como importantes
en la delimitación de grupos de especies afines
dentro de Justicia (Ariza-Espinar, 1971:
1988: Lester & Ezcurra, 1991).
Graham,
—
HISTORIA TAXONÓMICA Y DELIMITACIÓN DEL
GÉNERO
El género Justicia fue propuesto por Linné en
753. Un siglo después Nees, en su tratamiento de
las Acantáceas a nivel mundial (1847b). redelimitó
el género excluyendo más de 600 especies que se
habían descripto bajo el nombre de Justicia, re-
duciéndolo a solo 12 especies de Asia y África. Por
otro lado creó varios géneros nuevos afines a Jus-
ticia para ubicar un gran número de especies del
Nuevo Mundo recientemente descubiertas. Ben-
tham (1876) y Lindau (1895) ampliaron el concepto
de Justicia de Nees y redujeron a la sinonimia mu-
chos de los géneros creados por este último autor.
Durante este siglo, los estudios taxonómicos de Jus-
ticia han seguido estas dos tendencias opuestas a
través del tiempo, ya sea reconociendo un número
grande de géneros afines a Justicia s. str. (por ej.,
Rizzini, 1949, 1951;
tando una definición amplia que incluye a la ma-
Bremekamp, 1969), o adop-
yoría de éstos como sinónimos (por ej.. Leonard,
1951-1958). El problema de los límites del género
ha sido discutido ampliamente en épocas recientes
(por ej., Gibson, 1972; McDade, 1982; Ezcurra,
1988; Graham, 1988).
Actualmente la mayorfa de los autores tratan a
Justicia en su sentido más amplio siguiendo la pro-
puesta de Graham (1988) (por ej., Wasshausen,
1992: Profice, 1993; Daniel,
1995). Este concepto amplio ha resultado en esti-
1995; Kameyama,
maciones de 600—700 especies para todo el género
1988; McDade &
Daniel, 1998). Sin embargo, no se sabe si la gran
en los últimos años (Graham.
heterogeneidad de números cromosómicos que han
sido citados para el género (Daniel, 2000) estarían
reflejando evolución diversificadora a nivel cromo-
sómico, o características polifiléticas del mismo
(Daniel, 1993; Daniel et al., 1984; Piovano & Ber-
nardello, 1991). Análisis moleculares preliminares
(McDade & Daniel, 1998) indicarían que Justicia
s. lat. conforma un linaje separado del resto de la
tribu Justicieae, y que la gran diversificación mor-
fológica y la alta tasa de especiación en este género
se ha dado concomitantemente con muy poco cam-
bio a nivel molecular en los loci analizados hasta
el momento, lo que apoyaría la primera hipótesis.
Actualmente se está trabajando para dilucidar las
relaciones filogenéticas entre los taxa afines y per-
tenecientes a Justicia utilizando caracteres mole-
culares, morfológicos y cromosómicos, y un número
mayor de representantes especialmente del Viejo
Mundo (McDade et al.. 0)
Siguiendo el concepto de Graham (1988) el gé-
nero Justicia comprende especies caracterizadas
por corolas bilabiadas con surco estilar en la parte
posterior interna, presencia de 2 estambres exsertos
bajo el labio posterior, con 2 (raro 1) tecas, ausen-
cia de estaminodios, polen subprolado a perprolado
2- o 3(4)-porado o colporado, y cápsulas 4-semi-
nadas. Este criterio de delimitación del género es
el que se sigue en este trabajo.
MATERIALES Y MÉTODOS
El estudio se realizó sobre la base de material
de los principales herbarios de Paraguay y Argen-
tina, y de varios herbarios extranjeros (BM, CORD,
CTES, E, FCQ, G, K, LIL, LP, M, MO, NY, P, SI,
US, W (abreviaturas según Holmgren et al., 1990);
Apéndices 1, 2). Los ejemplares citados se selec-
cionaron como para ser representativos de la dis-
tribución y diversidad de cada especie en la región.
La identificación de las especies se hizo por com-
paración con material tipo o fotografías del mismo,
lo mismo que la resolución de los sinónimos no-
menclaturales. Las fotografías de ejemplares tipo
de la colección del Field Museum de Chicago, se
designan con la abreviatura "fot. Е y el número
Volume 89, Number 2
2002
Ezcurra 227
Justicia en Sudamérica Austral
de la colección fotográfica. Este estudio incluvó la
revisión de más de 220 ejemplares tipo.
En la mayorfa de los casos de nombres creados
por Nees (1847a. 1847b). este autor no designó ho-
lotipo sino que citó varios ejemplares representa-
tivos que deben considerarse sintipos. Debido a
que no pude consultar todos los sintipos de cada
nombre por estar estos ejemplares distribuidos en
varios herbarios europeos, en este trabajo general-
mente no designé lectotipos de entre estos sintipos.
Nees (1847a. 1847b) tampoco designó claramente
una variedad tipo cuando estableció varios taxa in-
fraespecíficos para una especie, sino que las desig-
nó con letras del alfabeto griego. Por las razones
ya expuestas previamente (Ezcurra. 1993€). consi-
dero en esos casos a la variedad alfa como la va-
riedad tipo de cada especie. En los casos en que
esle autor no designara una variedad alfa pero des-
eribiera variedades con letras subsecuentes (beta.
gama, etc.). considero a la descripta en la descrip-
ción general de la especie como la variedad tipo.
Otro problema que surgió en cuanto a la tipifi-
cación se relaciona con los ejemplares de Sellow.
Muchas de las especies descriptas por Nees
(1847a. 1847b) que se tratan en este trabajo fueron
descriptas sobre la base de material de Sellow del
sur de Brasil que estaba depositado en el herbario
de Berlín que fue destruido durante la segunda
guerra mundial. Existen. duplicados de varios de
estos ejemplares en Kew, los que se reconocen por
los datos “Herb. Reg. Berolinense. Brasilia. Sellow
legit." que figuran en las etiquetas. Como en gen-
eral estos son los tinicos datos que figuran en sus
etiquetas (falta información sobre la localidad y nú-
mero de colección). no se puede saber con segu-
ridad si realmente son duplicados de los holotipos
que estaban en Berlín que utilizó Nees para sus
descripciones. Pero cuando en estos ejemplares fi-
gura el nombre de la especie manuserito por Nees.
considero que muy posiblemente sean duplicados
del
este trabajo los cito como "probables?
mismo material que Nees vio en Berlín. y en
isotipos о
isosintipos.
El estudio de la distribución geográfica у de los
requerimientos ecológicos de las especies se hizo
sobre la base de datos del material de herbario
complementados con observaciones a campo de la
тауогѓа de las especies tratadas. La cita de una
especie para otros países siempre implica que he
visto material de esos países.
Los caracteres morfológicos se analizaron en for-
ma comparada en todas las especies presentes en
el área. agrupándose las mismas mediante búsque-
da de correlaciones que considero señalan posibles
afinidades filogenéticas. Muchas de las especies de
área en estudio eran muy poco conocidas y no ha-
bían sido tenidas en cuenta al elaborar el sistema
de clasificación vigente (Graham. 1988).
MORFOLOGÍA
La diversidad morfológica del género ha sido tra-
tada por Graham (1988). Aquí se discuten princi-
palmente los caracteres con valor taxonómico para
diferenciar las especies presentes en Argentina y
Paraguay. y se realizan observaciones en espectal
sobre las especies que no fueron incluidas en el
trabajo de Graham (1988) para ubicarlas dentro del
esquema de clasificación de esta autora (Tabla 1).
Irquitectura de la inflorescencia. Los tipos de
inflorescencia de las Acantáceas en general, inclu-
vendo varias especies de Justicia. han sido tratados
en detalle por Sell (1969, 1976). Graham (1988)
utiliza un sistema más simple de clasificación de
inflorescencias. que es el que se sigue en este tra-
—
bajo. Las especies de Sudamérica austral, coinci-
diendo en general con las del Nuevo Mundo (Gra-
ham. 1988) poseen (1) inflorescencias compuestas
(espigas de espigas, o panojas de espigas) (dibo-
trioides o tribotrioides según Sell. 1969). como en
J. corumbensis, J. jujuyensts, J. oblonga. J. oranen-
sis y J. saltensis, o (2) inflorescencias simples (es-
pigas simples) (monobotrioides segtin Sell. 1969).
como en J. axillaris, J. chacoënsis, J. cuspidulata.
J. gilliesii, J. hunzikeri, J. lilloana. J. phyllocalyx
y algunas formas de J. pectoralis y J. saltensis. o
(3) flores solitarias, como en J. riojana y J. twee-
diana y en algunas formas de J. axillaris. Las es-
pigas de Justicia muchas veces son secundifloras.
como en J. glaziovii, J. goudotii, J. laevilinguis y
J. saltensis. A veces en una especie se combinan
inflorescencia simples y compuestas, como las es-
pigas y racimos de espigas que se encuentran en
J. saltensis y J. xylosteides. Las flores solitarias y
las inflorescencias simples probablemente deriven
de inflorescencias compuestas (Sell, 1976). Además
de este tipo de condensación, en la evolución del
género también pudo haber habido elaboración de
inflorescencias simples a compuestas por combi-
nación de espigas axilares con espigas terminales
(Graham. 1988).
Morfología de la corola. Las corolas de Justicia
son marcadamente cigomorfas. Típicamente están
compuestas por un tubo basal cilíndrico, una gar-
каша más o menos tubulosa o campanulada api-
calmente dividida en dos labios, un labio superior
angosto, erecto. más o menos convexo y bidentado,
v un labio inferior ancho. más o menos patente y
trilobado. En las especies de Sudamérica austral
Annals of the
Missouri Botanical Garden
Table 1. Especies de Justicia presentes en Sudamérica austral, sección a la que pertenecen, probable tipo de
polinización, clase de ornamentación de la superficie seminal
Graham, 1988: A, tuberculada, verrugosa o pusticulada;
—
В, pilosa o papiloso-pilosa; С, lisa a muy levemente rugosa), y provincia fitogeográfica en la que se encuentran. (Con
asterisco* las
conoce su semilla madura).
especies que no habian sido classificadas por Graham, 1988, con signo de pregunta ? las que no se
Especie Sección Polinización Semilla Fitogeogr.
l. Justicia aequilabris Orthota Ornitofilia A Cerrado
2. Justicia axillaris* Dianthera. Ойна ct. Saglorithys Melitofilia A Paranaense
3. Justicia baenitzii Incierta Ornitofilia ? Yungas
4. Justicia brasiliana Plagiacanthus Ornitofilia C Paranaense
5. Justicia carnea Cyrthanthera Ornitofilia ? Paranaense
6. Justicia chacoénsis* Orthotactus Ornitofilia A Chaquefia
7. Justicia comata Incierta Melitofilia A Chaquefia
8. Justicia corumbensis* — Melitofilia C Chaquefia
9, Justicia cuspidulata* anther ra Saglorithys Melitofilia Cerrado
10. Justicia dumetorum сан Ornitofilia C Chaquefia
11. Justicia floribunda Plagiac — Ornitofilia C Paranaense
12. Justicia gilliesii* Sarotheca Melitofilia A Chaqueña
13. Justicia glaziovii Sarotheca Melitofilia A Cerrado
14. Justicia glutinosa Sarotheca Melitofilia A Chaquefia
15. Justicia goudotii Chaetothylax Lepidopterofilia B Chaqueña
16. Justicia hassleri Leucolomc Lepidopterofilia B Paranaense
17. Justicia hunzikeri* Orthotatus Melitofilia A Chaquefia
18. Justicia jujuyensis* Plagiacanthus Melitofilia C Yungas
19. Justicia kuntzei Sarotheca Melitofilia / Yungas
20. Justicia laevilinguis Dianthera subsect. Dianthera Melitofilia C Paranaense
21. Justicia lilloana* Incierta Melitofilia С Chaquefia
22. Justicia lilloi Simonisia Melitofilia C Chaquefia
23. Justicia lythroides Chae — Melitofilia B Paranaense
24. Justicia mandoni Simo Ornitofilia A Yung
25. Justicia oblonga* — Melitofilia B Paranaense
26. Justicia oranensis* Sarotheca Ornitofilia A Yungas
27. Justicia pectoralis Sarotheca Melitofilia A Chaquefia
28. Justicia phyllocalyx Simonisia Melitofilia ? Cerrado
29, Justicia polygaloides Dianthera subsect. Saglorith ys Melitofilia A Paranaense
30. Justicia ramulosa Chaetothylax rnitofilia B Paranaense
31. Justicia deg a* Plagiacanthus Melitofilia C Chaquefia
32. Justicia rusbyi Simonisia Melitofilia C Cerr. y Par.
33. Justicia saltensis* Sarotheca Melitofilia B Chaquefia
ЗА. Justicia squarrosa Simonisia Lepidopterofilia C Chaquefia
35. Justicia tocantina Chaetothylax Ornitofilia B Chaqueña
36. Justicia tweediana е, Melitofilia C Chaquefia
37. Justicia xylosteoides* Orthot Ornitofilia A Chaquefia
38. Justicia yhuensis a — Strobiloglossa Melitofilia B Cerrado
esta estructura básica está ampliamente diversifi-
cada, probablemente en respuesta a la presión se-
lectiva de diferentes tipos de polinizadores durante
la evolución del género. Esta gran diversidad mor-
fológica de las corolas, similar a la descripta para
las especies de Ruellia de Sudamérica austral (Ez-
curra, 1993c), comprende principalmente tres tipos
(Fig. 1,
El primer tipo se caracteriza por las corolas de
Tabla
color blanco o lila con el tubo, garganta y labios
proporcionalmente cortos y anchos, generalmente
con manchas y marcas venosas en la fauce, pro-
bablemente polinizadas por insectos del grupo de
las abejas (flores melitófilas, Fig. 1A, B). Este es
el grupo más numeroso en Sudamérica austral y es
muy variable en tamaño, ya que existen especies
con flores de este tipo de unos pocos mm de lon-
gitud (por ej., Justicia comata, J. glaziovii, J. poly-
galoides) hasta de varios centímetros de longitud
(por ej., Justicia lilloi, J. rusbyi).
El segundo tipo se caracteriza por las corolas de
color rosa, rojo o morado, grandes, con el tubo, gar-
Volume 89, Number 2 229
Ezcurra
2002 Justicia en Sudamérica Austral
Figura 1
lo role as ornit
C. D,
J. squarrosa.
— tweediana. jujuyensts.
J.
le ма rófilas: —E. J. жиш ой. —F
ganta y labios proporcionalmente angostos y largos,
aparentemente adaptadas a la polinización por pi-
caflores (flores ornitófilas, IC. D). En general
este tipo de corola mide de 2.5-5 ст de longitud.
Fig.
y también está muy representado en la región (por
| J. brasiliana, J. dumetorum.,
j-. Justicia baenitzii,
xylos-
J. floribunda. J.
teoides), lo que se relaciona con la importancia de
mandoni, J. ramulosa. J.
la polinización por picaflores en América del Sur
tropical y subtropical (Gentry. 1982)
Un tercer tipo se caracteriza por las corolas de
color rosado, lilacino o blanco, con el tubo y gar-
ganta Muy estrechos у el labio anterior extendido.
aparentemente adaptadas a polinización por mari-
posas (flores lepidopterófilas, Fig. IE. К) (por ej.,
Justicia goudotii y J. squarrosa de flores lilacinas,
y J. hassleri de flores blancas).
Muchas veces los distintos tipos de flores están
representados dentro de una misma sección de Jus-
ticia. Esto sugiere que, al igual que en Ruellia (Ez-
curra, 19936), la morfología de la corola de Justicia
pr
"A
Tipos de corolas de las especies Na Justicia presentes en Sudamérica austral.
J. | C.
B
E
y
Bn
F
( orolas me а Шах.
. B,
іаѕ. — . mandoni. — 1). as oranensis. E. "ui ds
(Esque Maro RA à partir de ilustraciones de Ezcurra. 19932.
ha resultado un carácter evolutivamente muy plás-
tico ante la presión selectiva de diferentes tipos de
flor
grupo.
polinizadores. De esta manera la forma de la
parece haberse diversificado mucho en el
incluso dentro de diferentes linajes, diferencián-
dose entre especies muy relacionadas en otros as-
pectos (como por ejemplo la ornitófila J. ramulosa
y la psicofila J. goudotii de Chaetothylax),
| general sea un carácter poco
a sect.
lo que hace que et
importante en la delimitación de grupos de espe-
cies relacionadas.
Morfología del polen. La llamativa diversidad
del polen de las Acantáceas en general y de Jus-
licia en particular ha sido descripto por Lindau
(1893, 1895), Raj (1961, 1973). Petriella (1968
Graham (1988), у Daniel (1998). Justicia se
teriza típicamente por el polen 2(—1) porado o 2(4)
Carac-
colporado, prolado, dorsiventralmente aplanado
cuando biaperturado, con el mesocolpio reticula-
damente esculturado, y la apertura con 1-3 hileras
230
Annals of the
Missouri Botanical Garden
de ínsulas o esculturas sexinosas a cada lado del
poro. Sin embargo, existen muchas variaciones de
1988).
Parte de la diversidad morfológica del polen de las
este esquema dentro del género (Graham,
especies de Justicia del sur de Sudamérica se de-
scribe en Petriella (1968) у Wasshausen y Ezcurra
(1997).
damente del resto del género por presentar toda la
Justicia phyllocalyx se diferencia marca-
superficie cubierta de ínsulas ordenadas en líneas
sub-paralelas al eje longitudinal (Wasshausen «
Ezcurra, 1997).
se asemeje a especies del género Potkilacanthus,
La cobertura de fnsulas hace que
un género muy afín a Justicia cuyas estrechas re-
laciones deberían estudiarse (Daniel, 1998).
Forma y superficie de la semilla. Como fue su-
gerido por Ariza-Espinar (1971), ciertos caracteres
de la morfología de las semillas han resultado muy
útiles para el reconocimiento de taxones infrage-
nérica en Justicia, en especial el grado de compre-
sión, el margen, y la superficie (Graham, 1988;
Lester & Ezcurra, 1991). Recientemente se ha pu-
blicado un estudio con microscopio electrónico de
barrido de semillas de doce especies de Argentina
(Peixoto, 1998). En el presente trabajo se presentan
microfotografías con MEB de otras nueve especies
del sur de Sudamérica (Figs. 2-4) que no fueron
tratadas por Peixoto (1998), y se dan las caracte-
rísticas macromorfológicas de las semillas de todas
as especies en las descripciones generales. Estas
características han sido utilizadas para clasificar o
confirmar la clasificación infragenérica previa de
2
as especies aquf tratadas (Tabla 1). La morfologia
de las semillas coincide con la descripta por Gra-
ham (1988) para las secciones en las que se ubican
las especies.
Las especies de Sudamérica austral poseen en
su mayoría semillas comprimidas, más o menos bi-
convexas y lenticulares, como J. cuspidulata (Fig.
2C), J. hassleri (Vig. 2E). J. lythroides (Fig. ЗС). J.
pectoralis (Vig. 3E). J. saltensis (Fig. АС) v J. xyloes-
toides (Fig. 4E). Algunas semillas comprimidas son
plano-convexas y tienen un reborde grueso en el
Fig.
A). J. brasiliana, y J. jujuyensis de la sección Pla-
—
margen del lado plano, como J. corumbensis
giacanthus. Una minoría de las especies de esta
región presentan semillas esféricas, como J. lilloi
(Fig. ЗА), J. rusbyi (Fig. ФА), J. phyllocalyx y J.
squarrosa, pertenecientes a la sección Simonisia.
La superficie en general varía entre las tres cla-
ses principales descriptas por Graham (1988) para
el género: (A) rugulosa, pusticulada, verrugosa о
tuberculada, (B) pubescente (pilosa o papiloso-pi-
losa), y (C) más o menos lisa (Tabla 1). Ejemplos
de superficie verrugosa a tuberculada (A) se en-
cuentran en J. cuspidulata (Fig. 2D) y J. axillaris
de la sect. Dianthera subsect. Saglorithis, J.
ob pa (Fig. 4F) de la sect. Orthotactus, y J.
comata de posición incierta. Ejemplos de superficie
pubescente (B) con pelos cortos aparecen en J.
hassleri (Fig. 2F) de la sect. Leucoloma, con pelos
más o menos gloquidiados en J. lythroides (Fig.
3D). J. oblonga y demás especies de la sect. Chae-
tothylax, y con pelos levemente capitados en J. pec-
toralis (Fig. 3F). En J. saltensis (Fig. 4D) de la sect.
Sarotheca los pelos son capitados y gloquidiados.
Superficies lisas (C) aparecen en J. laevilinguis de
la sect. Dianthera subsect. Dianthera y J. lilloana
de posición incierta, más o menos lisa en las es-
pecies de la sect. Plagiacanthus como J. corum-
bensis (Fig. 2B). v lisa y lustrosa en las especies de
la sect. Simonisia, como J. lilloi (Fig. 3B) v J. rusb-
1 (Fig. 4B).
Cromosomas. Se han estudiado los cromosomas
de varias especies de Justicia (Grant, 1955; Pio-
vano & Bernardello, 1991; Daniel et al., 1984,
1990; Daniel & Chuang, 1993; Daniel, 2000). El
número básico del género parece ser x = 7 y el
número más común n = 14. Sin embargo, no está
claro si la gran diversidad de los números (л = 7,
9-18, 20, 22-25, 27-29, 31 y 34) encontrados en
Justicia s.l. sugiere una evolución diversificadora o
un origen polifilético del género (Daniel et al.,
1984). Aunque se han encontrado algunas coinci-
dencias preliminares entre la clasificación infra-
genérica en secciones de Graham (1988) y los nú-
meros cromosómicos de las especies, está claro que
Justicia necesita todavía más estudio a nivel cro-
mosómico (Daniel, 2000). Los pocos recuentos pu-
blicados de especies de Argentina y Paraguay han
dado n = 14 o 2n = 28 (J. brasiliana, J. gilliesii,
J. oranensis, J. squarrosa y J. tweediana) excepto
en J. xylosteoides, 2n = 32 (Piovano & Bernardello,
199]
Distribución geográfica. Justicia es un género
muy diversificado en sus requerimientos ecológi-
cos. En Paraguay está representado en todos los
ambientes de las provincias fitogeográficas en que
se divide el país: del Cerrado, Paranaense y Cha-
quefia, у en Argentina también se encuentra en las
provincias de las Yungas, de la Prepuna, del Monte,
del Espinal y Pampeana (Cabrera & Willink, 1980)
(Tabla 1)
Especies xerófilas típicas de la provincia Cha-
queña son, por ej., Justicia squarrosa y J. xylos-
teoides, que aparecen exclusivamente en ambientes
semiáridos de la porción occidental de la región,
mientras que especies hidrófilas típicas de la pro-
J. brasiliana y J.
vincia Paranaense son, por ej..
Volume 89, Number 2 Ezcurra 231
2002 Justicia en Sudamérica Austral
Figura 2. Tint de semillas de las especies de Justicia presentes en Sudamérica austral. A, В. J. corumbensis
(Bernardi 20142). — specto general de la cara cóncava con reborde. —B. Detalle de la superficie iu C, D;
cuspidulata (Se hinini ) Bordas 20534). —C. Aspecto ge * өр —D. Detalle de la superficie verrugosa. E. F, J. ice ri
(Zardini y Velázquez 9943). —E. Aspecto general. —F. Detalle de la superficie cortamente verrugoso- —
carnea, que solamente se encuentran en bosques arenosos sometidos a incendios periódicos (por ej..
húmedos del noreste de Argentina y Paraguay ori- J. axillaris), o de bosques en galería de la región
ental. Existe un pequeño grupo de especies de Jus- chaqueña (J. corumbensis).
ticia endémicas de Sudamérica austral (J. gilliesii, No existe relación entre la taxonomía infrage-
J. hunzikert, J. riojana. J. lilloana). aunque la ma- nérica y la distribución geográfica. ya que dentro
voría en general tienen una distribución que excede de una misma sección puede haber especies de to-
los límites de la región. Muchas además son carac- dos los ambientes. Aparentemente el área extensa
lerísticas de condiciones locales y edáficas parti- y la diversidad de ambientes de la región subtro-
culares, como las especies de ambientes anegados — pical y tropical de América del Sur habrían per-
(por ej. J. laevilinguis), o de campos abiertos y mitido la evolución y diversificación de varios li-
232
Annals of the
Missouri Botanical Garden
Figura 3. Tipos de semillas de las especies de Justicia presentes en Sudamérica austral. A, B. J. lilloi (Mereles
2718). —A. Aspecto general de la forma esférica.
—B. Detalle de la superficie lisa. C, D, J. lythroides (Degen 1522).
—C. Aspecto general. —D. Detalle de la superficie pilosa, con pelos gloquidiados. E, F, J. pectoralis (Schinini y
Palacios 25788). —E. Aspecto general. —F. Detalle de la superficie pilosa, con pelos levemente capitados.
najes distintos de Justicia, muchos de ellos
representados por las numerosas especies presentes
en la gran heterogeneidad del territorio.
TAXONOMÍA
El análisis morfológico permitió confirmar la po-
sición taxonómica de varias de las especies pre-
sentes en la región, y proponer la ubicación de
otras 14 (marcadas con asterisco en la tabla) que
no habían sido tenidas en cuenta dentro del esque-
ma clasificatorio propuesto por Graham (1988) (Ta-
bla 1
rece ser robusto, ya que permite acomodar estas
—
. El esquema clasificatorio de esta autora pa-
especies.
En Sudamérica austral están representadas ocho
de las nueve secciones del género que existen en
el Nuevo Mundo: Justicia L. sect. Chaetothylax
(Nees) V. A. W. Graham, sect. Cyrtanthera (Nees)
V. A. W. Graham, sect. Dianthera (L.) V. A. W.
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
дм
4 Wes ДТ
~
Figura 4. Tipos de semillas de las es
2 | |
- Aspecto general de la forma esférica.
—D. Detalle a superficie |
F.
Aspecto general, ===
de |
Graham subsect. Dianthera. sect. Dianthera (L.) V.
\. W. Graham subsect. Saglorithys (Rizzini) V. A.
W. Graham, sect. Dianthera (L.) V. A. W. Graham
subsect. Strobiloglossa (Oersted) V. A. W. Graham.
sect. Leucoloma V. A. W. Graham, sect. Orthotactus
W. Graham, sect. Plagiacanthus (Nees)
. Graham, sect. Sarotheca (Nees) Bentham.
sect. Simonisia (Nees) V. А. W. Graham. El hecho
de que la región subtropical de América del Sur
sea un centro de diversidad tan importante para el
есте de Justicia presentes en Sudan
B.D
osa
Detalle de li
SA ee:
0 > A , 4
y 2 Ve Wey „ЕТ
э? T 29, 11
atl AM
al M.
NJ єс. а 2]
чў —
51
( vee
rusbyi (Arbo
rica austral. A, В. J.
isa. С, D, J. saltensis (Duré 428).
etalle de la superficie li ч
vylosteoides (Krapovickas э
sd
. соп pelos capitados. E. F,
(superficie verrugosa.
género sugiere una larga historia evolutiva en la
región y un posible origen sudamericano para el
mismo.
a siguiente. sinonimia incluye solamente los
nombres genéricos que han sido utilizados para es-
pecies presentes en Sudamérica austral. Para una
sinonimia más completa, véase Graham (1988) у
Daniel (1995). En esta lista se da el nombre de la
especie tipo de cada nombre genérico y su sinó-
nimo en Justicia. Los sinónimos homotípicos se
Annals of the
Missouri Botanical Garden
marcan con el símbolo =, y los basados en tipos
diferentes con el símbolo =. En Apéndice 3 se
presenta un índice de nombres científicos y de
nombres vulgares.
Justicia L., Sp. Pl. I: 1753, nom. cons. prop.
TIPO: (lectotipo, designado por Hitchcock en
Hitchcock & Green, 1930) Justicia hyssopifo-
ia L.
‚ 1753.
oe L., Sp. РІ. TIPO: Dianthera ame-
1 L. " Justicia americana (1. a e
— P Miller, Gard. Dict. Abr. Ed. + 754. TIPO:
Adhatoda zeylanesium Miller |= A adhatoda
Asiat. Rar. 3: 77, 112.
~
бии — Nees, in Wall., РІ.
1832. TIPO: mohos opia beyrichii Nees.
Beloperone Nees, in Wall., Pl. Asiat. Rar. 3: 76. 1832.
ТРО: (le ectotipo, designado por Graham, 1988) Be-
loperone amherstiae Nees |= Justicia brasiliana
Roth].
Leptostachya Nees, in Wall., Pl.
1832. TIPO: (lee totipo,
1988) — virgate
Asiat. Rar. 3: 76, 105.
designado por Graham,
ı [= Justicia virgata
Nees) T. s.].
dis a s i i indi. Nat. Syst. Ed. 2, . 1836.
TIPO: (lec — de signado por л 196€
)
Акуніна origanoic p ^s Nees [= Justicia a
(Nees) Griseb., non Vahl].
Ethesia Raf., Fl. Tellur. 4: 63. 1838, non Raf. 1837. TIPO:
thesia carnea (Lindl.) Raf. [= Justicia carnea
b
ana Raf., Fl. Tellur. 4: 61. 1838. TIPO: i ce
pectoralis (Jacq.) Raf. [= Justicia pec toralis Ra
MEME Nees, in Martius ras. 9: 847.
TIPO: (lectotipo, aquí Жашай) Orthotactus veno-
SUS Nees [= Justicia aequilabris (Nees) —
Heinzelia Nees, in Martius, Bras. 9: 153. . TIPO
(lectotipo, designado por Graham, ped Te lia
lythroides Nees |= Justicia lythroides (Nees) Gra-
ham].
rs Nees, in Martius, Fl. Bras. 9: 153.
"i ): Chaetoth ylax toc antinus Nees [= Justicia to-
A.W
~
-
Ww
Ww
-—
a (Nees) V um
C ae Ne es, in Martius, FL js 9: 99, 1847. TIPO:
(lectotipo, designado por Leonard, 1951-1958)
— magnifica Nees |= Justicia carnea
ndl. |.
Nees, in Martius, Fl. Bras. 9: 107. 1847.
Leonard,
Justicia serico-
Seric vara is
TIP
—
—
^
MT
л
(lectotipo, designado por -=
— E
o
his V. А. W. Graham].
—— Nees, in к tius, Fl. Bras. 9: 113. 1847. TIPO:
(lectotipo, — por Graham, 1988) Sarotheca
elegans Nees |= Justicia glutinosa (Bremek.) V. A.
.G сапап],
Seric — rigida Nees |=
оаа — . in Moric., Pl. Nouv. Amér.: 156. 1847.
ТІР — lepida Nees |= Justicia lepida
( е; Wassh. +
Fellium О. Kuntz „ Rev. Gen. 2: 486. 1891, non S. Kurz
'inónimo nomenclatural de Justicia L. 1753
* isc totipificación). TIPO: Justicia Farc gen L.
Chaetochlamys Lindau, Bull. Herb. Boiss. 3: 490. 1895.
PIPO: (lectotipo, designado por Graham, 1988) Che-
tochlamys mac rosiphon Lindau [= Justicia macrosi-
phon (L indau) V. Graham |.
Cyphisia Rizzini, Rev, Bios; Biol. 6: 521. 1946. TIPO:
‘YP hisia venusta. Rizzini [= Justicia venusta (Rizzi-
ni) V. A. W. Graham].
Аал ше — . Kon. Ned. Akad. We-
tensch., Afd. e de Sect. 45: 55. 1948.
TIPO: poen ae comata (L.) Bremek. [=
Justicia comata L..].
ME a pon Arch. Jard. Bot. Rio de Janeiro 8:
. 1948. TIPO: — paniculatum Rizzini
islicia л. Lindat
Acelica Rizzini Arch. Jard. Bot. Ro de Janeiro 9: 55.
1949. TIPO: Acelica cydoniifolia (Nees) Rizzini
sada en Adhatoda cydoniifolia Nees |= Justici
cydoniifolia (Nees) L —
dec Vie Arch. Jard. tio de Janeiro 9: 56.
. TIPO: Pupilla isis (Nees) Rizzini ba-
rid en Leptostachya heterophylla Nees |= Justicia
(Nees) Lindau |.
stris Rizzini, Dusenia 3: 189. 1952. TIPO: Thalestris
pile er iie Rizzini [= Justicia comata L.].
Thales
Hierbas perennes, sufrátices o arbustos erectos,
procumbentes o ascendentes, glabros o pubescen-
les, con cistolitos. Hojas opuestas, sésiles o pecio-
ladas, generalmente enteras, raramente crenadas o
crenuladas. Inflorescencia terminal o axilar, las flo-
res sésiles o cortamente pediceladas, con dos brac-
téolas, solitarias, fasciculadas o cimosas, en las axi-
las de brácteas, éstas frecuentemente dispuestas en
espigas o lirsos, que a su vez a veces se agrupan
en racimos o panojas; brácteas y bractéolas varia-
das. Cáliz profundamente 4- o 5-partido, los 4 o 5
segmentos generalmente angostos y subiguales. Co-
rola marcadamente cigomorfa, roja, morada, lilaci-
na o blanca, raras veces amarilla o naranja, el tubo
más o menos angosto a campanulado, derecho o
incurvado, con surco estilar en la parte posterior, a
veces ampliado en la garganta y con marcas colo-
readas en la fauce, del lado anterior, el limbo bi-
labiado, el labio posterior interno en la preflora-
ción, generalmente angosto, erecto o incurvado, a
veces cóncavo, con el ápice entero, bífido o leve-
mente bilobado, y el labio anterior más ancho, más
o menos patente o recurvado, profundamente a le-
vemente trilobado. Estambres 2, los filamentos ge-
neralmente insertos abajo. cerca, o por encima de
la mitad del tubo, filiformes o algo dilatados en la
base; anteras exertas, bitecas, las tecas general-
mente oblongas, a veces levemente curvadas o re-
niformes, discretas, superpueslas a cercanamente
paralelas en el conectivo, a veces la inferior redu-
cida, múticas, o una o ambas agudas, y apendicu-
ladas o espolonadas en la base; estaminodios au-
sentes. Polen prolado, 2- o 3- o 4-porado o
colporado, diversamente ornamentado, con la exina
Disco nectarífero anular,
reticulada. cupulado o
ciatiforme, entero o sinuado-lobado. Estilo filifor-
Volume 89, Number 2 Ezcurra 235
2002 Justicia en Sudamérica Austral
me, exerto, diminutamente bilobado en el ápice. — damérica, donde probablemente habiten cerca «
"
Ovulos 2 en cada lóculo. Cápsula de contorno 300 especies. Algunas de Sudamérica austral se
oblongo. elíptico u obovado, con la porción basal cultivan como ornamentales por sus flores lamati-
sólida y lateralmente comprimida, y la porción su- vas, como Justicia carnea (= Jacobinia pohliana.
"vara de la justicia”) y Justicia floribunda (= Ja-
cobinia pauciflora o Justicia rizzinit, “banderita es-
pañola”) (Parodi & Dimitri, 1980). aunque existen
perior cilíndrica. ovoide o subesférica, con los sep-
tos con retináculos. Semillas 4 o menos por aborto,
generalmente comprimidas o lenticulares, a veces
subglobosas. lisas o variadamente ornamentadas. . . : =
чуе аз, Sas РЕА л! rnamentadas varias otras nativas de Sudamérica austral con valor
rugosas, verrucosds o pilosas: retináculos CUPVOS V E | z d | ке 5
*.— ornamental que merecerían cultivarse, como J. ora-
agudos o truncados. 2.
"e : nensis y J. brasiliana. Algunas también tienen im-
El genero Justicia comprende más de 600 es- portane la económica como forraje ras de emergen-
pecies ampliamente distribuidas en ambos hemis- cia, en épocas de escasez en gramíneas, como
ferios. especialmente en los trópicos, extendiéndose Justicia tweediana (= J. campestris, “alfalfilla”) у
también a las regiones templadas del mundo. Pre- J. gilliesii (= J. echegarayi, “albahaca de vaca”)
Su- (Burkart, 1943).
senta un centro de diversidad importante e
CLAVE ARTIFICIAL PARA LAS ESPECIES DE JUSTICIA DE SUDAMÉRICA AUSTRAL
Flores de más de 2.5 em long.
xS
E
2a. s lilacinas, azules d blancas.
3a. Hojas axilantes de las flores o inflorescencias truncadas o cordadas J. lilloi
3b. Hojas axilantes de las flores o inflorescencias obtusas a redondeadas J. rusbyi
2b. Flores rosadas. rojas o mora
a. Flores en de isos tirsos terminales: corolas de 5-6 cm long J. carnea
tb. Flores en inflorescencias variadas. nunca en tirsos densos: уи menores.
За. Espigas con brácteas ovadas, elípticas u obovadas de más de З mm lat.
ба. Brácteas de 7-9 mm lat: cáliz 5-partido — — J. aequilabris
Ob. Brácteas de 3—5 mm lat.: cáliz 4 o 5-partido.
7a. Cáliz 4- — flores rojo-moradas J. ramulosa
7b. Cáliz 5-partido, flores rojo-anaranjadas J. oranensis
5b. ES con brácteas más angostas.
Flores con corola angosta, de 2.5-3 em long. por 2-4 mm diam.
Oa. Espigas densas; cáliz d-partido J. tocantina
Ob. Espigas laxas; cáliz 5-partido J. baenitzii
8b. Flores con corola más larga v ancha. Cáliz siempre 5-partido.
10a. Espigas secundifloras.
lojas subsésiles, de 0.7-1.5 em lat; arbustos xerófilos. Flores con labio
posterior de 1—1.5 em long. — J. x Mosteoides
11b. s pecioladas de 1.2-2.5 em lat.: balas o subarbustos higrófilos.
Flores con labio posterior de 1.52 cm long. m J. mandoni
10b.
— bilaterales o inflorescencias de otro tipo.
a. Flores solitarias o inflorescencias de espigas fasciculadas mucho más cortas
que las hojas axilantes.
13a. Inflorescencias axilares multifloras de flores rojas: brácteas espatu-
ladas u oblance lola a lineales J. brasiliana
13b. Espigas axilares paucifloras o utellus de flores con el ápice am
rillo: brácteas triangular-lanceoladas J foribunda
12b. Inflorese 'encias en espigas o racimos de espigas, más eh argas que las hojas
la. Brácteas angostas, de са. | mm lat, que ennegrecen al secarse, Flo-
res rojas J. chacoénsis
П. Brácteas más anchas. de más de 2 mm lat.. verdes. con el margen
blanquecino y hialino. Flores moradas а А dumetorum
Ib. Flores de menos de 2.5 em long.
15a. Corolas de 3-6 mm long. J. comata
15b. Corolas de más de 6 mm long.
l6a. Cáliz 5-partic
Іта. Inflorescencias generalmente terminales (a veces axilares también presentes).
é
ha
18a. Espigas capituliformes. condensadas en los extremos de las ramas. Brácteas y brac-
téolas largamente ciliadas. Semillas subesféricas —— J. squarrosa
236 Annals of the
Missouri Botanical Garden
18b. Espigas alargadas. Brácteas y bractéolas no largamente ciliadas. Semillas aplanadas
J. laevilinguis
17b. Inflorescencias generalmente axilares.
19a. Hojas pecioladas, de 2.5-5(-7) em lat.; plantas higrófilas.
20a. Inflorescencias pedunculadas mas largas que las —
21а. Corola de menos de 1 — ип long J. glaziovii
21b. Corola de más de 1.4 cm long.
22a. Flores blancas con mancha lila en la fauce |
Flores lilacinas con mancha blanca en la fauce
20b. e "las к — longitud o más cortas que las |
23a. Flo
J. corumbensis
J. kuntzei
a. Flores de ca. 1.5 em long. con segmentos del cáliz de menos de | em long.
| | ‚ Jujuyensis
23b. Flores de ca. 2 em long. con segmentos del cáliz de más de | ст long.
19b. Paje de menos de 2.5 cm lat.;
J. phy ое alyx
plantas xerófilas
Flores solitarias en las axilas de las айз superiores
25a. Cáliz ac — con los segmentos „йн. — lanceoladas a linea-
res, — menos de 0.3 em lat.
N
Р Е Е Е J. riojana
25b. Cáliz c Vi segmentos libres; hojas lanceoladas a ovadas, de 0.5-1.5
(2. 5) « em lat.
24b. — 's re nidis en espigas con brác ‘teas diferenciadas.
26a. nillas tuberculadas.
27a. Hojas de 0.3-1.5 cm lat., cortamente pecioladas J. hunzikeri
і Hojas de 1.8-2.5 cm lat., largamente pecioladas J. gilliesii
26b. Semillas lisas | - J. lilloana
lob. Cáliz 4-partido.
2
a. Flores solitarias en las axilas de I:
J. tweediana
Q7
as la is superiores o de brácteas similares a las hojas
9 J. axillaris
28b. — 's en inflorescencias ; multifloras con brácteas marc adame nte diferenciadas, mucho me-
no
292. — ncias generalmente de espigas simples.
30a. Hojas angostas, de menos de | cm lat.
30b. Hojas de más de | em lat.
3la. Hojas con lámina ovada a orbicular, de 2.5-5.5 em long
31b.
J. polygaloides
cuspidulata
Hojas con lámina angostamente ovada, ovado-lanc olada o lanceolada, de
1-7 cm long.
32a. Espigas de 4—6 cm long.; corolas de ca. 0.8 em long. Hojas de 1-2
em lat., generalmente pubescentes J. pectoralis
32b. Espigas de hasta 12 ст long.; corolas de 0.8-1.2 em long. Hojas de
1.5-2.5 em lat., generalmente glabras . yhuensis
29b. Inflorescencias compuestas de espigas agrupadas en racimos (raramente espigas sim-
ples en las axilas de las hojas).
33a. Infloresce ncias con los ejes y las brácteas glandulosos.
a. — ovada o elíptica de 7-15 «
nm long
Ww
—
m long.; espigas densas; brácteas de 3—
. glutinosa
34b. Lámina ovals de 3-10 cm long.; espigas laxas; brácteas de 2-3 mm long.
ч J. saltensis
33b. Inflore scene las no glandulosas.
cáliz de ca. 0.6 cm long.; anteras con la teca apical perfecta y la basal
ob "o ta; reducida a un leve engrosamiento sobre el filamento J. lythroides
35b. Cáliz de más de 0.6 em long.: anteras con la teca apical perfecta y la basal
reducida pero no obsoleta.
36a. Brácteas, bractéolas y segmentos del cáliz albo-marginados. Corola
de ca. 2.5 em long. . hassleri
36b. Васе, bractéolas у segmentos de | cáliz totalmente verdes. Comla
de 1-2 em long.
37a. Corola de 1.5-2 cm., con tubo largo de 1—1.5 em long.
J. goudotii
37b. Corola de ca. 1 em, con tubo corto de 0.5-0.6 cm long
J. oblonga
l. Justicia aequilabris (Nees) Lindau, in Engler (Nees) Nees, in DC., Prodr. 11: 358. 1847. Ja-
& Prantl, Nat. Pflanzenfam. 4(3b): 350. 1895
cobinia aequilabris (Nees) Lindau, Bull. Herb.
Orthotactus aequilabris Nees, in Martius, Fl.
Boissier 3: 486. 1895. TIPO: Brasil, s. loc.,
Bras. 9: 134. 1847. Amphiscopia aequilabris Sellow 174 (holotipo, B destr., fot. F 8902)
Volume 89, Number 2
2002
Ezcurra 237
Justicia en Sudamérica Austral
Orthotactus strobilaceus Nees, in Martius, Fl. Bras. 9: 133.
847. Amphiscopia strobilacea (Nees) Nees, in DC..
Prodr. 1 l: 358. 1847. Justicia strobilacea (Nees) Lin-
i gler И риа. Nat. РИ; 1(3b):
350. 1895. “TIP Brasil. — fl. San Fran-
cisco, in sylvi des Antonio el Alegre сеа Mar-
tius — м! ) «ж Blanchet 2575 (sintipo, Wi;
isosintipos, BM!,
кз arnolttianus E . in M: artius, Fl. Bras. 9: 132.
TIPO: Brasil. — Blanchet 2575 ex herb.
D si sert (isotipo. e
Orthotactus venosus Nees, in Martius, Fl.
1847. si s venosa (Nee
Prodr. 11: . 1847. Justicia renosa (Nees
Kew Bull. ps (ay 617. 1988. TIPO: p E
ranhao: Pohl 18-49 (holotipo. d. fot.
iu velutina Lindau. Bull. Herb.
189 ‚ише nebutina (Lindau) Lindau ex Hassler.
6 1917
Mi. PI. Hass TIPO: Brasil. Glaziou
anzenfam.
Bras. 9: 133.
Nees, in DC.
¢) Graham.
Rio Ma-
Eu 3: 487,
076 бейти, 1 !
R!
ои к “а Lind Bull. Herb. Boissier 3:
1806. cla durian (Lindau) Gra-
. 43 (4): € 19 TIPO: Bolivia.
: 000-800 m. و san. (holotipo. В
F 0900!: !
de str. fot. isolipo, NY
Sufrútice de 0.3—0.5 m alt., con tallos subcilin-
dricos, pubérulos. Hojas con pecíolo de 0.7-1.5 em
long., pubérulo, y lámina ov е a anchamente elíp-
пса. de 4-9 em long. por 2.5-5 em lat.. cuneada у
decurrente en la base, — u obtusa en el ápice.
levemente pubérula en el haz, velutino pubérula en
е | envés. cm
Inflorescencias en espigas de 5—7
subsésiles y alternas en las axilas de las ho-
le
cm — у 0.7-0.9 em lat., con la venación
long..
imbricadas. ovadas.
jas superiores; brácteas
1.3-1.7
marcada. muy levemente pubérulas, ciliadas: brac-
téolas lanceoladas de ca. | em long. y 2 mm lat..
elabras, ciliadas. Cáliz profundamente 5-partido.
con los segmentos lanceolados, de ca. 5 mm long..
generalmente glabros. Corola roja de 3—3.5 cm
long.. con tubo basal angostamente cilíndrico. de
2.5 ст long. y hasta 4 mm diám.. el labio posterior
bidenticulado. de | em long. por 3.5 mm lat..
anterior trilobado, de ea. 1 em long., con los lóbulos
de 6-7 mm long. Estambres exertos con las anteras
debajo del ápice del labio superior, las tecas a dis-
tinta altura, ambas de ca. 2 mm long.. müticas.
Cápsula claviforme de ca. 1 em long., con semillas
aplanadas y (según Nees, 1847a) tuberculado-ve-
rrugosas de ca. 2 mm long.
Ilustraciones. Nees, 1847a: tab. 21 (sub Ori-
hotactus venosus).
Este de Bo-
livia, Brasil y noreste de Paraguay, en la región de
Distribución, hábitat y fenología.
Amambay. Habita en bordes de bosque y florece en
primavera, de octubre a noviembre.
Justicia aequilabris se caracteriza por las espigas
axilares con brácteas imbricadas, ovadas, anchas y
papiráceas. y las flores de corola roja generalmente
de más de 3 cm long. Es una especie variable y de
distribución amplia (ver tipos citados en sinoni-
mia), aunque aparece raramente v está relativa-
mente poco coleccionada, como lo demuestra la fal-
ta de ejemplares de Bolivia y Brasil en los
herbarios. Los nombres que se consideran sinóni-
mos presentan ejemplares tipo cuyas pocas diferen-
clas parecen formar parte de la variabilidad mor-
fol6gic ‘a de esla espec ie,
Material seleccionado. PARAGUAY. Amambay: In
dumetis Ponta Рога. Sierra de Amambay. Rojas 10929 (G).
: Zwischen Rio Apa und Rio Aquidaban.
E, G, P)
Concepción
Centurion, Григ 1103 (BM.
in Engler &
1895.
Justicia axillaris (Nees) Lindau.
Prantl, Nat. МЭЬ: 350.
Pflanzenfam.
Rhytiglossa axillaris Nees, in Martius, Fl.
Bras. 9: 122. 1847. TIPO: Brasil. Rio Grande
Barbara da Eneruc il-
D destr.. fot.
do Sul: fl. Jacuy ad S.
hada. Sellow s.n.
9011).
(holotipo.
Rhvtiglossa campestris Nees, in Martius, Fl. Bras. 9: 118.
1847. Justicia campestris (Nees) Lindau, in Engler
& Prantl. Nal Рт nfi am. 4 (3b): 350. 1895. non
Justicia campestris Griseb., Goell. Abh. 10: 225.
874. TIPO: Brasil. Rio — do Sul: Porto Alle-
1. (h m B destr.. fot. F
!: probables latinos. EL
Hierba pequeña, rizomatosa, de hasta 15(—30)
em alt., con tallos erectos o algo decumbentes. sub-
cilíndricos. delicados, con dos líneas longitudinales
у opuestas de pubescencia fina y curva en cada
entrenudo, surgiendo de las axilas de las hojas. 3
raíces engrosadas. Hojas subsésiles a cortamente
pec ¡oladas. las inferiores orbiculares a ovadas, de
2. y 1-1.5
superiores ovadas o angostamente ovadas а elípti-
5 ст long. em lat., redondeadas. las
cas (raramente orbiculares), de 1-2 em long. y 0.3—
em lat.. agudas. todas glabras о hirsuto-pubéru-
las. cilioladas en el margen. Flores bibracteoladas
sésiles en las axilas de las hojas superiores o de
brácteas algo más angostas y pubescentes, poco di-
lerenciadas, a veces formando espigas terminales:
bractéolas lanceoladas, de 0.7 a 1.5 em long. y ca.
l mm lat.. pubérulas. садах. Cáliz profundamente
I-partido con los segmentos lineales, de 0.5-1 cm
long., generalmente ciliados. Corola lila. pálido-li-
lacina a blanquecina de 1.4-1.7 em long.. pubé-
rula. Labios de aproximadamente la misma longitud
que el tubo basal, el posterior brevemente biden-
tado, el anterior profundamente trilobado, con los
lóbulos de 5-6 mm long. y 3 mm lat.. el lóbulo
central más ancho y venoso-rugoso en la fauce. Es-
tambres con las tecas divergentes. y estilo filiforme
238
Annals of the
Missouri Botanical Garden
con estigma bilobado. Cápsula claviforme, de 7 mm
long. y 2 mm diám., glabra. Semillas algo aplana-
8 ) B go а
das, de 1.5-2 mm diám., amarillentas, tuberculado-
verrugosas; retináculos de 2 mm longitud.
Ilustraciones. | Dawson, 1979: 573.
Distribución, hábitat y fenología. Sur de Brasil,
Uruguay, este de Argentina y Paraguay oriental, en
campos abiertos arenosos o pedregosos. Florece en
primavera y verano, de octubre a marzo.
Justicia axillaris y Justicia campestris parecen
corresponder a formas con distinta morfologia foliar
y diferente cantidad de pubescencia de una misma
especie polimórfica característica de campos abier-
tos, por lo que el último nombre se trata como si-
nónimo del primero. Por su cáliz 4-partido, brác-
teas angostas y semillas tuberculadas, se clasifica
en Justicia sect. Dianthera subsect. Saglorithys (Ta-
bla 1). Justicia reitzii Leonard (Tipo: Brasil. Santa
11 ao Norte de Abe-
lardo Luz, campo, 500—600 m, 25 Dic. 1956, Smith
y Reitz 9230, holotipo, US!) es una especie rara del
Catarina: Abelardo Luz,
sur de Brasil muy afín a J. axillaris, que se dife-
rencia por las flores siempre solitarias en las axilas
de las hojas superiores y el tamaño generalmente
menor y la textura más coriácea de sus láminas
foliares. A pesar de que recientemente se citó como
1999), se consi-
dera que las relaciones entre ambas deberían es-
sinónimo de J. axillaris (Ezcurra,
tablecerse sobre la base del estudio de más mate-
rial del sur de Brasil.
Material seleccionado. ARGENTINA, Corrientes:
Concepción, — “е 906 (CTES); Ituzaingó, Ea. El
Plata, Meyer 6414 (LIL); Santo Tomé, Gdor. Virasoro, Es-
tabl. Timbauva, a más o menos 9 km de 14, monte
próximo al casco, Romanczuk 188 (SI); San Miguel, 21 km
al S de Loreto, Schinini 8327 (CTES); Ituzaingó, Playa-
dito, 20 km W de Apóstoles, Schinini 21794 (CTES): San
Roque, M. P. Mansilla, Ybarrola 2866 (LIL). Entre Ríos:
Concordia, Ayuí, alrededores Hotel Salto Grande, flor lila,
campos pedregosos, Troncoso 1158 (SD; Federación, Salto
Grande, laderas pedregosas, Troncoso 27231 (S1); La |
La Paz a
siones: Capital, Posadas.
—
me
x
Feliciano, Troncoso y Bac — 3337 (SD.
ab: Burkart 14171 (S
* rt 14280 (SI); fon:
Candelaria, Ruiz
=
es, campi
pland, campos, Jörgensen 69 AL);
ШШ 4876 (LIL). РАК, stan-
Elena, Pira Pyta, Schinini y Caballero Marmori
27168 (С TES). ( — Pr. Caaguazti in campo :
Hassler 9271 (G). Canindeyú: in regione Yerbalium de
Maracayü, Paraguaria euro austral, Hassler 5108 (BM, С).
Cordillera: in fluminis Y-acá, il in campo pr. Valenzue la,
sicco,
62 (BM, G, NY):
»)
‹
"Plantae Pos den А ' Hassler 1907 (G
3. Justicia baenitzii (H. Winkler) C. Ezcurra,
Bol. Soc. Argent. Bot. 25(3—4): 348. 1988. Be-
loperone baenitzii H. Winkl., in Fedde Repert.
7: 113. 1909. TIPO: Bolivia. San
Carlos bei Mapiri, in Walden, 750 m, Ago.
1907, Buchtien 1409 (holotipo, B destr., fot. F
8922!; isotipo, US!).
Nov. Spec
Be lope: rone — Rusby, Mem. New York Bot. Gard.
7: 307. TIPO: Bolivia. Huachi, 1800 ft., 13
Ago. D. "his 550 (holotipo, NY!; isotipo, Us).
Sufrátice perenne, a veces decumbente o apo-
vante, de cerca de 1 m alt., con ramas cilíndricas,
levemente 4-suleadas, geniculadas por encima de
los nudos, glabras o pubérulas. Hojas con pecíolo
de 1-3 em long. y lámina ovada o anchamente elíp-
tica de 8-20 em long. por 3-7 cm lat., general-
mente glabra, acuminada en el ápice, cuneada y
algo decurrente en la base, lisa a ondulado-crenada
en el margen, con las venas principales pubérulas
y prominentes en el envés. Flores en espigas uni-
aterales laxas agrupadas en racimos axilares y ter-
minales que superan ampliamente a las hojas en
longitud. Raquis pubérulo; brácteas de las espigas
lanceolado-lineales de 2—4 mm long.. glanduloso-
pubérulas; bractéolas en las bases de las flores li-
neales, de 2 mm long., pubérulas. Cáliz de 3—4 mm
long. con los 5 segmentos triangulares, agudos, de
2.5—3 mm long., glanduloso-pubérulos. Corola roja
de 2.5-2.9 cm long., con tubo de cerca de 1.5 cm
long.. angosto, la garganta de 4 mm diám., los la-
bios de 1.5 em long., el posterior levemente biden-
tado, el anterior trilobado y levemente reticulado-
venoso en la fauce. Estambres con filamentos de 1
v anteras con tecas desiguales, una de 1.5
cm long.
mm long., la otra de más de 2 mm long., la superior
mútica, la inferior apendiculada en la base. Cáp-
sula inmadura claviforme, de ca. 1 em long. por 2
mm diám., pubérula, con la mitad inferior angosta
y aplanada y la mitad superior inflada y apiculada.
Semillas maduras desconocidas.
Ilustraciones. Ezcurra, 1993a: 342.
Distribución, hábitat y fenología. Perú (fide
Brako & Zarucchi, 1993) y Bolivia hasta el Norte
a Argentina en selvas de la provincia fitogeo-
—
de
gráfica de las Yungas. En Argentina ha sido encon-
trada únicamente en el extremo NW de la provincia
de Salta, y en el departamento de Ledesma, en la
provincia de Jujuy. Es frecuente en el sotobosque
de la selva entre los 600 y 1500 m s. mar, y florece
en invierno y primavera, de julio a octubre.
Justicia baenitzii se asemeja al material tipo de
Jacobinia tenuistachys Rusby (Tipo: Bolivia. Be-
tween Guanai and Tipuani, 1892, Bang 1441, ho-
Volume 89, Number 2 239
2002 Justicia en Sudamérica Austral
lotipo, МҮ!; isotipo. E!) en la morfología de la in- simples o compuestas, de hasta 3—5 em long.. más
florescencia y el tamaño de las hojas. Sin embargo
el tipo de J. tenusitachys no presenta flores desa-
rrolladas y además tiene hojas sésiles, por lo que
no parece tratase de la misma especie.
Material seleccionado.
Martín, PN Calile
ARGENTINA. Jujuy: San
egua, Río Las Piedras, Judica y aed
Santa Victoria, PN Baritú. Nacientes del
de я fini а de Y Үе er rvoorst y Cuezzo 7725 C (LIL):
— fon а C — 3 km from Aguas Blancas to Fin-
i 150 m. — 1955 (K): Orán. Vado
e Willink 133 (LH
4. Justicia brasiliana Roth. Sp. Pl. Nov: 17.
1821. Beloperone brasiliana (Roth) Bremek..
Verh. Kon. Nederl. Akad. Wetensch. Amster-
dam. Afd. Naturk. sect. 2, 45(2): 52. 1948
TIPO: Orig. de Brasil. cult. en Europa (no vis-
Lo).
Justicia nodosa Hook.. Bot. Mag. 56: tab. 2914. 1829.
Dianthera — soot) Benth.. in Bentham &
дн ег, Gen. Pl. 2 (1): 1113. 1876. TIPO: Brasil.
ult. en Liverpool Bot. Gard. (holotipo, K!).
виле —— Nees. in Wallich, Pl. Asiat. Rar. 3:
102. 18 TIPO: "Ex horto ill. Amherstiae speci-
mina. a i. in Indam orientalem translata." in
herb. Wallich (fide Nees, 1847b) (no visto).
Beloperone amherstiae um var. debilis Nees,
Fl. Bras. 9: 139, TIPO: Brasil.
S... pr. prae de Mineiros,” Martius s.n. (sin-
tipo. MP: Porto Allegre, Sellow s.n. (sintipo, B destr.:
isosintip
in. Martius.
“In svlvaticis
po
Be рем — Nées var. — A es. in Mar-
tius. Fl. Bras. 9: 139. 18 TIPO: Brasil. Montis
Buta агау. Sellow 120 е В destr.: isotipos.
‚ K).
in Martius,
Sellow 145
Be ums amherstiae je es fo: ciliaris Nees,
Fl. Bras. E 139, TIPO: Brasil.
olotipo. B dest
Beloperone — Ne "es و sd Nees, in Mar-
tius, Fl. Bras. 9: 139. TIPO: Brasil. In the
woods of Hio Grande. — 752 (holatipa, K!).
Sericoeraphis — а Nees, in DC
1847. TIPO: Bri er
10lot "ipo, я fot. US
Jacobinia he Rizzini, Ted 5)6): . 5. 1975.
TIPO: Brasil. Rio de Janeiro: Bus p жын Riz-
zini y Ini ne 21 (holotipo. RB!).
361.
Tweedie s.n.
Prodr.
Rio Grande de Sul:
Arbusto o hierba sufruticosa robusta, de 1-2 m
alt.. con tallos erectos o apoyantes, cilíndricos. gla-
bros. Hojas cortamente pecioladas. con láminas
lanceoladas a angostamente ovadas. de (2.5—)4—
1018) em long. por (12)2-3(4) em lat..
periores menores. acuminadas en el ápice y obtusas
—
as su-
a cuneadas en la base, coriáceas, discolores. en-
teras o algo crenadas. glabras a muy levemente pu-
bérulas. Flores sésiles dispuestas en espigas densas
cortas que las hojas, cortamente pedunculadas у
opuestas en las axilas de las hojas superiores. Brác-
teas axilantes de las flores oblanceoladas a espa-
tuladas o lineales, de 8-12 mm long. por 1—2(—3)
mm lat.. una fértil y la otra estéril en cada nudo.
glabras a levemente pubescentes; bractéolas linea-
es a lanceoladas de ca. 5 mm long. por 0.5 mm
lat.. semejantes a las brácteas. Cáliz profundamente
>-partido, de 4-5 mm long., con los segmentos lan-
ceolados. de 3—4 mm long., agudos. Corola roja de
3-4.5 cm long.. con tubo de ca. 2.5 cm long.. es-
trecho. de hasta 5 mm diám. en la fauce. y labios
de 1-1.5 em long., el posterior erecto, angostamen-
te ovado, levemente bilobado, y el anterior patente.
trilobado. con los lóbulos de 3-5 mm long. por 2—
3 mm lat., redondeados. Estambres más cortos que
el labio posterior, con las tecas a distinta altura. de
Estilo fili-
y ovario glabro.
1-1.3
con la mitad inferior sólida y lateralmente
ca. 2 mm long.. la inferior sub-oblicua.
forme con estigma bilobado. Cáp-
sula claviforme castaño-brillante. de mm
long..
estrechada. y la superior subesférica. de 4 mm
diám. Semillas aplanadas de contorno orbicular, de
З mm «ат... lisas, pardo-claras, con un reborde
en su cara interna: retináculos de ca. 2 mm long.
Ilustraciones. Wasshausen y Smith. 1969: 96
(fig. 14A).
Sur de Brasil.
norte de Uruguay. noreste de Argentina y Paraguay
Distribución. hábitat y fenología.
oriental, desde el nivel del mar hasta los 600 m.
Es una especie higrófila que habita en claros y bor-
des de bosques húmedos y en selvas. frecuente-
mente en lugares alterados como márgenes de ríos
y bordes de picadas. Florece principalmente en ve-
rano y otoño. desde noviembre a mayo.
Justicia brasiliana se caracteriza por sus flores
angostas de color carmín de hasta 4.5 cm long. dis-
puestas en inflorescencias axilares densas, con
brácteas oblanceoladas a espatuladas o lineales. Es
una de las Acantáceas más frecuente en bosques
húmedos de Paraguay oriental y del sur de Brasil
y este de Argentina. y por sus flores llamativas. más
coleccionadas en la región. A pesar de haber sido
introducida en cultivo en Europa tempranamente
en el siglo XIX bajo los nombres de Justicia bra-
siliana y Justicia nodosa, actualmente no se registra
como planta cultivada en Argentina (Parodi & Di-
mitri, 1980)
ARGENTINA. Corrientes:
ue 'Tón de. \ы rada, Berón de Astrada, Паре, Ruiz Huidobro
Grande, — — 4249
desvío Km 395.8 FCNEA. Ruiz Hui-
Capital, Ruta 5, 19 mn del Triángulo,
Material seleccionado,
Santo Tomé, Cuay
ries an "Май,
dobro 4301 (LIL):
Annals of the
Missouri Botanical Garden
Schinini y Martínez Crovetto 12771 (SI); Mburucuyá, Ea.
Santa Ana, Schwarz 94 (LIL); Concepción, Río Santa Lu-
cía, Schwarz 9275 (LIL); CT, Dpto. Saladas, Ayo. Ambro-
sio, Schwarz 9394 (LIL). Chaco: CH, Resistencia, Mar-
garita Belén, Aguilar 566 (LIL); Libertad, ruta 16, a orillas
del Río Negro, — ee 32 (LIL); Libertad,
Popular, Schultz 16244 . Formosa: Mojón de
Krapovickas 1086 (SI); — Ruta 11 vieja, pasando
Arroyo F. C i |, Guaglianone 726 (SI); Pilco-
mayo, Puerto Ramos, Morel 7182 (LIL); — Ma
l Guayacán, Reales 94 (LIL).
Javier, costas Río Uruguay, Bertoni 2513
; Guaraní, Gob. Morales y Fracrán, —
Sáenz 29211 (S . Alem, 7 Km E de L. Alem, lot
Marufiak 678 (Li 1 " Oberá, Oberá, Villa Blanquita, е
нет 242 (LIL); San Martin, San Martín, De Marco 10654
C (LIL, M); 5 A, Schwarz САГУ
Iguazú, Ruta Nac. 101, 5 :
Colonia
Fierro.
Misiones: San
—
—
—
entre
an Ignacio, obraje EM
5 km del límite E del P. Nac..
Zuloaga 5270 (S1). PARAGUAY. Alte > Paraná: i in regione
а Alto den Fiebrig 6174 (BM, E. К, LIL, SI,
; Pr. Escuela Técnica Fore sx km 12 Puerto Stroess-
ner, Bernardi 18891 (BM, G, ), NY). Amambay: Par-
que Nacional Cerro Cora, al in del río Aquidabán cerca
del fortín — Brunner 1534 (G. MO, US); Pedro Juan
Caballero, Bernardi 19330 (С.
O, NY). a: Arroyo Yakare'i, along N side from
route 2, Zardini y Aguayo 10730 (MO, US): Tebicuary Mí.
in nemore collino Isla Alta, Be rnardi 18720 (6. MO, NY).
Caazapá: Tavai, 500 m d 3894 (CTES):
R. Pirapó, N of José Bassai, West dus 1). Canindeyú:
—
s Palos ad septentr.,
. Soria .
Casas y Molero
5855 (MO, NY): in e yerbalium de Maracayu. Pa-
raguaria euro-austra, Hassler 5207 (BM, G., K, P, W). Cen-
tral: I Assomption, dans les lieux ombragés. Balansa
2459 (K, P); fig Centralis, Hes lacus Ypacaray,
Hassler 11740 (BM, Е.С. K, LIL, MO, US). Concepción:
entre Paso Horqueta у unes ión, Krapovickas et al.
14227 (MO, US); Zw. Rio ja und Rio Aquidaban, Villa
Sana, Fiebrig 4662 (BM, E, K, M). Cordillera: in regione
lacus Ypacaray, Hassler 11105 (BM, K, MO, US); Montes,
San Bernardino, Hassler 1593 (С. Р, S1); on silvis pr. Cor-
dillera de Altos, Hassler 1579 (С. К, P. US); Paraguaria
ce entralis, in silva San Bern: silicio: Hassler 3016 (BM, G
. MO, Р, W). Guairá: Cordillera Ybytyruzú, Villa San
Pedro. — 1876 (MO); Road Melgarejo—Antena, 8
km f Melgarejo. Za rdini 11349 (MO, US): Tororo, ca-
mino à v rro Polilla, Degen 1032 (G); Tororo, Cerro Асат,
Soria 2588, 2637 (С); Villa-Rica. dans les bois, Balansa
2459a (P). Itapúa: Encarnación, Mboycaé, Bertoni 4024
LIL). Misiones: San Ignacio, Estancia Brusquetti, Mere-
les 1198 (G). аа Pilar, Curupayty, Meyer 1595(
(LIL); Villa Pilar, — 7876 (LIL).
e Acahay Massif, on rocks on SE peak. Zardini
et al. 12782 (MO, US); Cerro Mbatoví, in forest, Zardini
y Soria x (MO. US); Cerro Palacios, Soria 2242 (G);
Soria 2720 (G);
Escuela Agricola,
Ypoa, bosque wi borde ‘a
Parque Nac e Ybyci
quez 12094 (M
1189 (MO). Presidente Hayes: in regione
rioribus pps Pilcomayo, Rojas 664 (G):
Chaco, km . Mereles 3904 (US). San Porn 0 km
US); Loma
. —
el lago.
0); Dau.
N de San E тне d Krapovickas et al. 13€
Hoby, Woolston 193 (SD, Woolston 193C —
Central Paraguay, Morong 200a (ВМ. С. К. MO,
5. Justicia carnea Lindley, Bot. Reg. 17: t. 1397.
1831. Ethesia carnea (Lindl.) Raf.. Fl. Tellur.
4: 63. 1838. Jacobinia carnea (Lindl.) Nich-
olson, Ill. Dict. Gard. 2: 206. 1885. TIPO: Bra-
sil. Rio de Janeiro: enviado a Inglaterra e in-
ilustrado en Lindley
troducido en cultivo,
(1831).
irr id magnifica Nees, in Martius, Fl. = 9: 100,
1847. Jacobinia n s, Vilm.
1. 3) 1: 81
PME, (ed. 3) 1894. Jacobinia mag-
nifica (Nees) L — in & Prantl, Nat. Pflan-
zenfam. 4(3b): 351. TIPO: Brasil. B de Ja-
xeiro: Tijuca, — s.n. ш. V.
elo: — Nees var. minor Nees, in Martius,
Fl. Bra . 1847. TIPO: Brasil. Semidorio, Rie-
del s.n. — BR, fot. US); Rio de Janeiro,
Corcovado, — s.n. (sintipo, LE no visto).
Conant мег эр ees, in Martius, Fl.
17. Je abia — (Nees) Voss,
=
шо (ed. 3) 1: 810. 1894. Jacobinia ane
(Nees) L — in Engle r & Prantl, Nat. Pflanzenfam.
4(3b): 3 1895. TIPO: Brasil. S. est.: pr. Mideiros
el Entre 2 Morros, Pohl 196 (sintipo, W!); Minas
Gerais: in silvis primaevis ad Cabo d'Agosto et Villa
do Principe, Martius s.n. _(sintipo, M!); Rio Grande:
Porto Alegre, Serra de 7, 252
(sintipos, B destr.); Tres Irmáos, Sellou
— y; Praesidio eu Joao Baptista, d 254
(sintipo, В destr., fot. F 8915!, AURA: K!).
C ушш pohliana Nees var. in Martius,
01. 1847. Jain ohio (N ees) L.
lew. — ntes — l: 136. 1923. TIPO: Brasil.
* Raven (sintipo, M!); Ins.
Catarit HOC SMS s.n. (sintipo, LE no visto).
C йыга аала Nees ls velutina Nees, in Martius,
Fl. Bras. 9: 101. 18 . Jacobinia — (Nees)
Vilm. Blume — (ed. 3) 810. 1894.
Brasil. Rio de Janeiro: — ‘rs a fine rose
— s.n. (sintipo, K!); Argentina. Buenos
Aires: “one of the . . . ornaments of the gardens of
s Avs flowers all the year,” Tweedie s.n. (sintipo,
K!).
Orthotactus ciliatus Nees, in Martius, Fl.
1847. — Ep _ciliata (Nees)
Prodr. 59. 1847. TIPO: Brasil.
Monte С — e hnatt s.n.
(holotipo, M!).
=
Н. Ba
"Absque d o nala
Voss,
TIPO:
colour,”
Bras.
Nees, in
Rio de Janeiro:
in herb.
35.
-
A
Martius
Arbusto o hierba sufruticosa robusta de hasta 2
m alt., con ramas subtetrágonas, glabras o corta-
mente pubescentes. Hojas con pecíolo de 3-5 cm
long. y lámina ovada o elíptica de 10-25 cm long.
por 5—8 cm lat., obtusa y levemente acuminada en
el ápice, cuneada en la base, parcialmente decu-
rrente sobre el pecíolo, de textura herbácea, áspera,
con el margen irregularmente crenado, con las ve-
nas principales marcadamente prominentes y no-
torias en el envés, las secundarias arqueadas y pa-
ralelas, la nervadura en general pubescente. Flores
sésiles dispuestas en tirsos terminales densos y
bracteados, ovoides, de 10-20 cm long. y 5-8 cm
didm., bractéolas
floribundos. Brácteas folidceas;
Volume 89, Number 2
2002
Ezcurra 241
Justicia en Sudamérica Austral
lanceoladas, de aprox. 15 mm long. por 2 mm lat.,
obtusas, glabras y ciliadas. Cáliz profundamente 5-
partido, con los segmentos lanceolados, de 10 mm
long. por 2 mm lat., agudos. Corola rosada o roja,
de cerca de 5-6 cm long., con tubo de ca. З cm
long.. ensanchado a 3—4 mm diám. en la garganta,
y labios de ca. 2.5
bidentado, el anterior patente y trilobado. con los
lóbulos de 3 mm long. Estambres inclusos debajo
cm long., el posterior curvo y
del labio posterior, con filamentos de ca. 4 cn
long.. y anteras con tecas a distinta altura: la su-
perior separada por el conectivo y oblicua, de ca.
Fruto desconocido.
Ilustraciones. Nees, 1847a: tab. 14
Cyrtanthera magnifica). Wasshausen y
1969: 90, fig. 13B (sub Jacobinia carnea).
Sur de Brasil.
Habita
3 mm long. ambas.
(sub
Smith,
Distribución, hábitat y fenología.
noreste de Argentina y Paraguay oriental.
en bosques húmedos, en claros y bordes de pica-
das. frecuentemente en terrenos bajos y cerca de
cursos de agua. Florece principalmente de prima-
vera a otoño, entre noviembre y mayo.
Nombre vulgar. “Vara de la justicia” en Argen-
tina (Parodi € Dimitri, 1980), “Plume flower”
Nueva Zelandia (Webb, 1995), “Brazilian plume
en Estados Unidos (Daniel, 1995).
Justicia carnea es una especie del sur de Brasil
en
у regiones limítrofes cultivada como ornamental en
lugares de clima cálido de todo el mundo por sus
grandes inflorescencias de flores rojas o rosadas,
que aparecen durante todo el año, hasta en am-
bientes sombríos. Su cultivo puede haber extendido
su distribución natural original, ya que se ha asil-
vestrado en regiones donde ha sido introducida. Por
ejemplo ha sido citada para Ecuador y Colombia
(Leonard, 1951-1958: & Smith,
1969). y como introducida y naturalizada en Nueva
Zelandia (Webb, 1995: 154)
seleccionado. ARG
Wasshausen
Material QENTINA. Misiones:
Iguazú, Salto Iguazú, ши 789 (К, Sl): General Ma-
nuel Belgrano, reserva estricta de San Antonio, Múlgura
1891 (S1): San Pedro, serva Esmeralda, Deginani 1320
(SD: General Manuel DN San Antonio. 9 km de San
Antonio, Deginani 2080 (S1): San Javier, Arroyo Patino,
Schwarz 4176 a IL). EA UAY. Alto Parani: in re-
gione fluminis Alto Paraná, Fiebrig 5730 (BN LIL,
SI), Fiebrig 5826 (G, LIL, SD: procedente puerto — rloni.
Alto Paraná, culta Jardín puta o, Rojas 10688 (LIL).
6. Justicia chacoénsis Wasshausen & С. Ezcu-
rra, Candollea 52: 178. 1997. Beloperone ri-
paria S. Moore, Trans. Linn. Soc. London, Bot.
1: 432. 1895, non Justicia riparia Kameyama,
Bol. Bot. Univ. Paulo 14: 204. 1995.
TIPO: Brasil. Near Corumbá. Moore 1047 (ho-
lotipo, BM!: isotipo, B destr.. fot. F 89441).
Sáo
Arbusto o sufrútice ramoso de 0.80-1.50 m alt.,
con tallos glabros, ramificados, de corteza pálida.
Hojas con pecíolo de 0.5-1.5 cm long. y lámina
ovada a anchamente elíptica, de 6-13 cm long. y
2.56 cm lat.,
decurrente en la base, algo discolor, glabra en el
haz, velutino-pubérula y glabrescente en el envés.
Inflorescencias en espigas densas, de 3-8 cm long..
acuminada en el ápice, cuneada |
en las axilas de las hojas superiores, éstas a veces
reducidas y entonces las espigas reunidas en un
racimo apical denso, el conjunto oscureciéndose al
secarse; brácteas angostamente lanceoladas, de 1—
1.3 em long. por ca. 1 mm lat., largamente acu-
minadas, pubérulas, negras al secarse; bractéolas
lineales, de 3-5 mm, más angostas, pubérulas. Cá-
liz profundamente 5-partido, con los segmentos lan-
ceolados, de 4 mm long. por menos de | mm lat.,
agudos, pubérulos. Corola de ca. 4.5 cm long., roja,
con tubo angostamente cilíndrico de 2-2.5 em long.
y labios de 1.5-2 cm long.. el posterior angosta-
mente oblongo. agudo, el anterior de 0.8 em long.,
profundamente trilobado. con los lóbulos angosta-
mente oblongos, de 0.7-1 ст long. y З mm lat.,
redondeados. Estambres exertos casi hasta el ápice
del labio posterior, las anteras con las tecas mar-
cadamente superpuestas, de ca. 2 mm long., las
inferiores separadas de las superiores ca. 2 mm,
algo menores y basalmente apiculadas. Cápsulas
angostamente claviformes, de 1-1.3 em long. por
2-3 mm lat., con la base sólida y lateralmente es-
trechada en el tercio inferior, muy levemente pu-
bérulas. Semillas aplanadas, de 2 mm diám., ve-
rrucosas, oscuras; retináculos de ca. 1 mm long.
Ilustraciones. Figura 5.
Distribución, hábitat y fenología. Sudoeste de
Brasil, norte de Paraguay y regiones limítrofes de
Bolivia. Habita en bosques xerófilos abiertos, y flo-
rece en primavera y otoño, de septiembre a mayo.
Justicia chacoënsis se caracteriza por sus flores
rojas grandes, de cerca de 5 em long., en inflores-
cencias espiciformes con brácteas pequeñas y an-
Por sus caracte-
gostas que oscurecen al secarse.
rísticas ubicó en Justicia sect.
Orthotactus (Wasshausen & Ezcurra, 1997).
morfológicas se
Material seleccionado. PARAGUAY. Chaco: Agua
Dulce, Schinini y Bordas фе (GC. МО); —— lo
* renza, Pique Histórico, ca. 10 km SE de P. Lageren-
‚ Charpin y Ramella 21610 ( 3); Cerro León, Schinini y
Bordas 17794 (US); Cerro León, desde lomada al S (cam-
— hasta meseta бешш, Charpin y Ramella 21742
;); Fortín Gabino, Mendoza a Lagerenza, a 12 km. desvío
a TE Cué, Degen 3280 (FCQ); Mayor Pedro Lageren-
sque abierto de quebracho у samohá, Schinini у
Bordas 14999 (US); Palmar de las Islas a Ravelo, Cerro
san Miguel, Mereles y Ramella 2795 (ЕСО, G
242
Annals of the
Missouri Botanical Garden
/
1
сор
Figura 5. Justicia chacoénsis. —A. Rama con flores.
Flor con bráctea, cáliz, corola, estambres y estilo. —C.
j).
Corola abierta con androceo. Fortunato 3738 (G), Billiet y Jadin 3263 (G
4. Justicia comata (L.) Lamarck, Encycl. 1: 632.
1785. Dianthera comata L., Syst. Ed. 10: 850.
1759. — hya comata (L.) Nees, in DC.,
Prodr. 381. 1847. Ecbolium comatum (L.)
s — Can. Pl. 2: 487. 1891. Stetho-
ma comata (L.) Britton, Bot. Porto Rico 6: 218.
1925. Psac adoc — comatum (L.) Bre-
. Wetensch.. Afd.
1948. TIPO:
Jamaica. Brow ne s.n. (lectotipo, — por
Graham (1988), LINN, microficha BM!).
‚ Ned.
Tweede Sect. — 55.
— martiana Nees, in Martius, Fl. Bras. 9: 152
1847. TIPO: Brasil. Bahia: — — М!),
Salzmann s.n. (sintipo, B destr., ; isosin-
tipo, ВМ!); s. loc., Sellow s.n. тке B кн иен prob-
able isosintipo, BM!); Pará, Poeppig s.n. (sintipo. B
destr.). Cuba. S. loc.: Poeppig s. n. (sintipo, B destr.).
Leptostachya martiana Nees var. hisp artius,
d à jw . 9: 152. 1847, TII PO: Guyana. Británica.
urgh s.n. nibe, K no visto).
e martiana Nees var. macrophylla Nees, in
la Nees, in
Martius, Fl. Bras. 9: 152. 1847. TIPO: Guyana Bri-
lánica. Schomburgk s.n. (holotipo, К!; isotipo, E!
er nya parviflora Nees, in Martius, Fl. Bras. 9: 1 51.
18 TIPO: Brasil. Bahia: Ilheus, Махат
Vied s.n. (sintipo, BR no visto), — 125 (sintipo,
3 destr., probables isosintipos, K!, E!); Soteropolin,
Manche s.n. (sintipo, G no visto; isosintipo, M!);
;astelnovo, Riedel s.n. (sintipo, LE no visto).
ا acuminata Nees, in DC., Prodr. 11: 354.
1847. Justicia acuminata (Nees) Lindau, in Engler
& ет Nat. ae fa. A(3b): 351. 1895. TIPO:
Mexico. asco: marais de Teapa, Finden 1633 (lec-
totipo, Meiner por - Daniel — 5), K
Justicia Swartz, Prodr.: 14. 1788. ж Jamai-
са. Swartz s.n. (holotipo, Bu ів про, М!).
Thalestris nai Pau Dusenia 3(2): 190. 1952.
TIPO: Brasil. Paraná: Parque Nacional do Iguazú,
Duarte 1705 (holotipo, ier isotipo, NY!).
3
m
—
Hierba perenne, con largos rizomas horizontales
provistos de raíces adventicias en los nudos, y ta-
llos erectos de 30-80 cm alt.,
glabros o muy levemente pubérulos. Hojas opues-
cilíndricos, frágiles,
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
tas, sésiles, subsésiles o cortamente pecioladas, an-
gostamente ovadas, oblongas o lanceoladas, de 4—
(0.8)1-2 ecem lat..
redondeadas. cordadas o cuneadas en la base, fre-
8(-15) cem long. por
cuentemente decurrentes sobre el pecíolo, agudas.
levemente discolores. glabras o pubérulas. Flores
muy pequenas dispuestas en espigas secundifloras
laxas у delicadas que se agrupan formando. seu-
doverticilos en inflorescencias paniculiformes la-
xas, terminales o axilares, con los ejes filiformes.
elabros o levemente pubérulos, y con algunos pelos
elandulares. Brácteas axilantes de las flores lan-
ceoladas, de 0.8-1.2 mm long., generalmente gla-
bras: bractéolas algo más pequeñas. Cáliz profun-
damente 5-partido, con los segmentos lanceolados.
1-2.5 mm long. por menos de 0.5 mm lat., agu-
dos. glabros o levemente glanduloso-pubérulos. Co-
rola blanca, lilacina o violácea. de 3-6 mm long..
elabrescente. con el tubo de 2 mm long., la gar-
ganta de | mm diám., los labios de cerca de 2.5
mm long., el posterior emarginado, el anterior tri-
lobado. reticulado-rugoso en la fauce. con las mar-
cas moradas. los lóbulos de aprox. | mm long. por
| mm lat. Estambres con filamentos de ca. 2 mm
long.. y anteras con las tecas situadas a distinta
altura. divergentes, de ca. 0.4 mm long.. la inferior
algo menor, separadas por el conectivo. Cápsula
pubérula de 3-47) mm long.. con el tercio infe-
rior sólido y lateralmente comprimido, la porción
superior engrosada, de contorno oblongo. Semillas
1. aplanadas, orbiculares. de 1 mm diám.. dimi-
nutamente verrugosas о pusticuladas, pardas: reli-
náculos pequeños.
Ilustraciones. Ezcurra. 1993a: 352.
Distribución, hábitat y fenología. Ampliamenle
distribuida en América tropical desde el sur de
México hasta Bolivia. Paraguay y norte de Argen-
tina. En el sur de su distribución se encuentra en-
tre el nivel del mar y los 600 m de altitud. Habita
en suelos anegados de charcas. bañados, esteros y
pantanos. o en lugares muy húmedos y soleados de
bordes de bosques y selvas. También invade terre-
nos removidos húmedos y adopta características de
maleza. Florece casi todo el año, principalmente en
verano, de diciembre a febrero.
Justicia comata se caracteriza por las flores muy
pequeñas dispuestas en inflorescencias paniculifor-
mes tenues y laxas con ramificaciones verticiladas,
de aspecto graminoide, y no tiene afinidades claras.
Como varias otras especies higrófilas de ambientes
anegados, es muy variable morfológicamente y tie-
ne un área muy extendida, y eso ha dado lugar a
la creación de numerosos nombres que resultan en
su extensa sinonimia.
Material seleccionado. ARGENTINA, Chaco: Orillas
del Río Guaycurú, Meyer 2094 (LIL). Corrientes: Gen-
eral Paz, Lomas de Krapovickas y Cristóbal
11812 (SD: Paso de los libres. Paso Troncón. Ruiz Hui-
dobro 3714 (LIL); Alvear, Ea. Santa Ana.
guay, Schinini 16905 (LIL » San Martin, peyd; costa Río
Uruguay, Ybarrola 1891 (
mún en monte y esteros. a nsen 2832
a
Vallejos.
~
~
ph
e
en chacra з. Jorge nsen 3278 (Sl): Pirané. al S del pueblo
— Mor 605 (LIL): Pilcomayo, Puerto ~ Моге!
7190 ( у Jujuy: 5 h anta — Ruta Prov. 1. Laguna
i . Misiones: — Cata-
San Ignacio. Jardín
: i esa 5967 (LIL): Santa
Ana, — Rodr rigues 755 (К, LIL); San
1944
=
=
3
=
Ф
m эм
`5
z
м
=
=
>ч
Q
Yı
©
—
as
ее
Arroyo. ltacaruaré, camino vie jo. Zuloaga
turi 5591 (К. PAR jet »n
Fiebrig 1423 (K, G, M). Alto Dua: in regione fluminis
Alto Paraná, — 6081 (G, K, LIL, SI, — Rio Mon-
day. Bertoni 1378 (LIL). — Bella Vista, Río \ра.
Schinini y Bordas 20628 (C N
de Coronel Oviedo, Krapor ickas r 2 (CT S);
3 km S del destacamento, Rio
(СТЕ. MO). Puerto Há Enram: —
cano al Rio Paraguay, Se — y Bordas 1331-4 (CTES
MO. US): San Lorenzo, Campus sS ts
1208 (CTES, С); Villa a dans les mares, Balansa 2.108
(6. К. Concepción: г. ‚Ара. on Rio —
Villa Sana, Fiebrig 4688 "Ki lillera: ad ripam
lacus Ypacarat, Hassler — (BM. С. K, LIL. MO Ny.
P); Cercanías de San Bernardino, Mereles 3270 (G). Guai-
Schinini y Bordas 25242
: Común en el agua. Villa Rica. s.f..
NY. SI, US). Itapúa: Encarnación. Río
Colonia Encarnación. Montes 715
Isla ( fous Pe-
ene ron: 20 km
ES, l Tavaf,
Te E пагу. Soria 3303
Central: arroyo | cer-
га: С xw Independencia,
(CTES
3796 (L n MO.
Paraguay.
nes: Santiago, pU ча La Soledad,
dersen 5994 (K, MO, NY). Presidente Hayes: in regione
cursus inferioris — Pilcomayo, Rojas 362 (BM
К: Ruta trans- Chaco, km 219, Schinini y Palacios 25
K). San Pedro: Colonia Primavera. Woolston
770 (K, LIL, NY, SI), Woolston 1383 (K). S. dep.: Para
guaria Seple ee — 1907 (G. К): 38
Hassler 5985 (С. K, LIL, MO, NY. Р)
Jörgensen
. Misio-
G;
To
DL
8. — corumbensis (Lindau) Wasshausen &
Candollea 52: 178. 1997. Belope-
rone m Lindau. Bull. Herb.
2 5: . 1905. TIPO: Brasil. Mato Gros-
so: pr. — Abr. 1903, Malme 3029 (ho-
lotipo. B destr.. fot. F 392891).
. Ezcurra,
Boissier
SGE. 2. ws +
Hierba sufruticosa de 0.6-1 m alt, con tallos
subcilindricos, cuadrisulcados, pubérulos a gla-
bros. a veces con dos líneas longitudinales pilosas
B |
en su juventud. Hojas con pecíolo de 0.5-1 cm
long. y lámina ovada de 6-12(-18) cm long. por
2.5-5(-1) em lat., aguda y acuminada en el ápice.
a J
obtusa a redondeada en la base, decurrente sobre
el pecíolo, discolor, con cistolitos poco notorios en
su superficie, subcoriácea. generalmente glabra.
C
Flores sésiles dispuestas en espigas pedunc e Is.
Annals of the
Missouri Botanical Garden
A
Figura 6. Justicia corumbensis. —A. Rama n flores.
—C. Corola abierta con androceo. =).
estilo.
(SI).
éstas solitarias o de a pares en las axilas de las
hojas superiores. Espigas frecuentemente iguales o
más largas que las hojas, densas, con pedúnculos
de 2—4 cm long., y raquis a veces ramificado en la
base; brácteas florales espatuladas a obovadas, de
ca. 5 mm long. por 1.5-2.5 mm lat., acuminadas,
foliáceas, generalmente glabras, desiguales en cada
nudo, una fértil y la otra menor y estéril; bractéolas
oblanceoladas, de menos de 5 mm long. Cáliz 5-
partido, de 3—4 mm long., con los segmentos tri-
angulares, de ca. 2 mm long., agudos, glabros, ci-
liolados en el margen. Corola blanca con marca lila
en la fauce, de ca. 1.5 cm long., con tubo de 5-6
mm long. ensanchado en una garganta acampanada
de 4 mm diám., y labios de 5—7 mm long., el pos-
terior brevemente bidentado, erecto, cóncavo. de 4.
{/
oD
E
—RB. Flor con bráctea, bractéolas, cáliz, corola, estambres y
"stilo y estigma. Е. Cápsula abierta. Е. Semilla. Jörgensen 2337
~
el anterior de 7 mm lat., ampliamente tri-
mm lat.,
lobado, patente, con los lóbulos de ca. 2.5 em lat.,
redondeados. Estambres con filamentos de 4-6 mm
long. y anteras con las tecas a distinta altura, la
superior de 1.5 mm long., la inferior oblicua, algo
mayor y con un apéndice glandular en la base. Es-
tilo glabro; estigma redondeado y ovario glabro.
Cápsula glabra o pubérula de ca. 1 em long., con
la mitad inferior sólida y lateralmente comprimida,
y la superior subesférica y 4-seminada, de 3—4 mm
diám. Semillas aplanadas, de contorno orbicular, de
2-3 mm diám., lisas, pardo-anaranjadas, con un
reborde en su cara interna; retináculos de cerca de
2 mm longitud.
Ilustraciones. Figura 6.
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
Distribución, hábitat y fenología. Sudoeste de
Brasil,
oeste de Paraguay, principalmente en la región del
sudeste de Bolivia. noreste de Argentina y
Chaco. Habita a lo largo de los bosques ribereños
de los ríos de la región chaqueña. Florece en ve-
rano, de enero a abril.
Justicia corumbensis pertenece a la sección Pla-
giacanthus y es una especie muy afín a J. juyuyensts
. Ezcurra (Wasshausen € Ezcurra, 1997) con la
que suele confundirse en herbario. Ambas se ase-
mejan en el tamaño, color y forma de las brácteas
y flores. y en las características de los frutos. Jus-
ticia corumbensis se diferencia de J. jujuyensts por
las flores en espigas pedunculadas frecuentemente
más largas que las hojas axilantes, mientras que J.
jujuyensis posee inflorescencias axilares más cortas.
Justicia jujuyensis es característica de los bosques
pedemontanos de transición entre las provincias fi-
togeográficas Chaqueña y de las Yungas. mientras
que J. corumbensis generalmente habita en los
bosques en galería de la región del Chaco septen-
trional.
Material se ا А ARGENTINA. Chaco: Maipú.
ruinas km 75, ruta go 15846 (CTES); —
Schulz 3235 (CORD, t TES 1082: Pirané: R. 90, 15
km al S de Pirané. pr Ade 680 (SI); Común en A
monte. km 139 del F.C.. Jörgensen 2336 (LIL, SD, 2337
(SD: Patiño, гата
(СТЕУ): Matacos, Ing. Juárez. Morello s.n. herb., Se
18213 — PARAGUAY. Alto Paraguay:
ad meridiem Fuerte Olimpo.
guay, Bernardi 20457 (MO). Presidente — Estancia
M Wap a durs del río Siete Puntas, Degen y Mereles
3250 (FCO): Km 326 de Asunción, haud proc al ab Pozo
Colorado. — i 12 (G. MO, NY, US): Santa Elisa,
Hassler 2754 (BM. С. K, NY j
Bartolomé de las Casas,
©
. Justicia cuspidulata (Nees) Wasshausen. Mo-
Syst. Bot. Missouri Bot. Gard. 45: 1253.
1993. Rhytiglossa cuspidulata Nees, in DC.,
поют.
Prodr. 11: 348. 1847. TIPO: Perú. Mathews
3152bis (holotipo, К!; isotipos. sub Mathews
3152 BM!, B destr., fot. F 184.12).
Hierba sufruticosa en la base, rizomatosa, erecta,
de hasta 50 em alt., con tallos angostos, subcilín-
dricos, poco ramificadas, hirsuto-pubérulos. Hojas
de 2.5-
obtusa o aguda en el
subsésiles con lámina ovada a orbicular.
5.5 em long. у 1.8-3 cm lat.,
ápice, truncada a levemente cordada en la base.
— en el haz. y pubescente a pubérula en «
— 'encias en espigas terminales laxas
de 21 1 em long.. con el raquis glanduloso-pubes-
cente; brácteas triangulares de 34 mm long. рог
1.5 mm lat., pilosas: bractéolas de la misma lon-
gitud. algo más angostas. Cáliz profundamente 4-
partido, con los segmentos lineal-lanceolados, de
6-8 mm long., acuminados, pilosos. Corola lilacina
y 2
mm diam. en la fauce, y labios de 5 mm long., el
de ca. 8 mm long., con tubo de 4 mm long.
posterior oblongo. de 2 mm lat., bidentado, y el
anterior trilobado, con los lóbulos de 3 mm long.
por 1.5 mm lat., redondeados. Estambres exertos
hasta la mitad del labio superior, con las tecas su-
perpuestas, oblicuas, marcadamente separadas por
el conectivo. Cápsula de 8 mm long. por 2 mm de
grosor, glabra. Semillas aplanadas, de 1 mm diám..
diminutamente verrugoso-tuberculadas o pusticu-
ladas, pardo-oscuras; retináculos de menos de |
mm long.
Ilustraciones. Figura 7.
Distribución, hábitat y fenología. Originalmente
descripta para Perú, se la ha encontrado en el este
de Bolivia y en el noreste de Paraguay. en la región
del Amambay. Probable en Brasil. Habita en bordes
de bosque. Florece en primavera y verano, de agosto
a marzo,
El material de Paraguay que aquí se determina
con el nombre de Justicia cuspidulata pertenece a
Justicia sect. Dianthera subsect. Saglorithys, un
erupo de especies afines muy diversificado en Amé-
rica del Sur que se caracterizan por las inflores-
cencias en espigas simples y las flores medianas,
con cáliz 4-partido y semillas tuberculadas (Tabla
1). A este grupo también pertenecen por ejemplo
Justicia rusbyana Lindau (Bolivia. Yungas. Bang
370, isotipo, M), Justicia boliviana Rusby (Bolivia.
Vic. Cochabamba, Bang 1225, holotipo, NY!; iso-
tipos. K!. US), y la especie descripta como Rhyti-
glossa. piahuiensis Nees (Brasil. Piahui, Gardner
2290, holotipo, K!: isotipo, BM), especies muy afi-
nes a Justicia cuspidulata. La delimitación entre las
especies de este grupo no es clara, y eso hace es-
pecialmente difícil asignarle un nombre al material
de Paraguay. Por otro lado el tipo de Justicia cus-
pidulata, aunque morfológicamente muy similar al
material de Paraguay, es de Perú, y existen pocos
ejemplares que claramente parezcan pertenecer a
esta especie en el área intermedia entre este país
y Paraguay. Sin embargo. la flora del bosque tro-
pical estacionalmente seco presenta distribución
discontinua en Sudamérica, y existen varios otros
ejemplos de especies de esta flora que presentan
Ч
al., 2000). Por estas razones, antes de publicar un
este tipo de distribución disyunta (Pennington
~
nombre nuevo para el material de Paraguay, prefie-
ro identificarlo con el de Justicia cuspidulata, con
el que está indudablemente relacionado. El nombre
Justicia amamba yensis C. Ezcurra que utilicé para
esta especie en determinaciones en etiquetas de
herbario hace unos años nunca fue publicado.
246 Annals of the
Missouri Botanical Garden
M Zi duas
: — — ——
E
1 f x
x V.
seD
Figura 7. Justicia cuspidulata. —A. Rama con flores. —B. Flor con bráctea, cáliz, corola. estambres ye
2
=
C. Corola abierta con androceo. —D. Cápsula abierta sin semillas. —E. Semilla. Soria 2036 (G ). Schinini 20534 (С).
Volume 89, Number 2
2002
Ezcurra 247
Justicia en Sudamérica Austral
El tipo de Rhytiglossa cuspidulata (Mathews
3152bis) está mezclado con el de Rhytiglossa hoo-
keriana (Perú. Chachapoyas, Mathews 3152). una
especie con hojas más pequeñas y flores más an-
gostas, como queda claramente explicitado en su
descripción original, y los isotipos que se encuen-
tran en BM tienen la identificación invertida.
Material seleccionado. BOLIVIA. Santa Cruz: Bue-
cras, ears h 6973 (BM). Beni:
— Estación Biológica Beni, са. 1/2 hour by jen
o R. Curiraba. Solomon 14633 (M). PARAGL
ede Amambay. Ruta 3 y Río Aquidabán. Se ы
y Bordas 20534 (CTES, US): in regione cursus py rioris
fluminis or Hassler 8003 (BM, С. NY), 1 (BM, G,
MO, NY. Р); orillas del Río ш: ruta 3. Kra-
pávickas 15995 (С — Parque Nacional Cerro Cora, ori-
llas + — Soria. Zardini y Ortiz 2036 (С. FC Q. MO
re › Apa und Rio Aquidabán, he — 4406 (BM. 3
= P): Caballero Cué, Fiebrig 4473 (BM, E. G, K.F
du Торак. Cerro Tobatí, in nas 287 (CTES): *
batí, Sparre y Vervoorst 1412 (LIL).
navista, desmontes chac
10. Justicia dumetorum Morong, Ann. New
York Acad. Sei. 7: 193. 1893. TIPO: Paraguay.
Río Pilcomayo, 1888-1890, Morong 1538 (ho-
lotipo, NY!
CE N. E. Brown, Trans. Bot. Soc. Edinburgh
: 67. 1896. TIPO Río Pilcomayo
1890-189 l.
патот моз Lindau, Bull.
2, 73. 1905. TIP о
Santa f lisa, Fab. 1903. leg. Rojas, Н
lotipo, B безі, fot. Я я 1201:
LIL MO! NY! PL
Argentina.
x
т
=
Seed
BM! GL KL
isolipos,
Hierba sufruticosa de 0.8-1.5 m alt., con ramas
erectas, cilíndricas. longitudinalmente estriadas, en-
grosadas por encima de los nudos. Hojas con pecíolo
de 0.5-2 em long. v lámina ovada a lanceolada de
4-12 cm long. por 1-3 em lat., aguda y acuminada
en el ápice, cuneada o redondeada en la base. ge-
neralmente glabra, con cistolitos pequeños densa-
mente esparcidos en su superficie. Espigas simples.
densas, apicales y en las axilas de las hojas supe-
riores, de 3-10 ст long.. a veces agrupadas en ra-
cimos foliosos en el extremo de las ramas. Brácteas
lanceoladas, opuestas, de 1—1.5 em long. por 1.5-2
mm lat., glanduloso-pubérulas, con el margen blan-
quecino y Майпо, cada una con una flor sésil en su
axila y dos bractéolas de 7-9 mm long. en la base.
Flores con el eáliz de la misma longitud que las
bractéolas. profundamente 5-partido, los segmentos
de 6-10 mm long. por 1.5 mm lat.. glabros y albo-
marginados. Corola de color morado de 34. cm long.
con tubo de 2-2.5 cm long.. poco ensanchado en la
fauce. y labios de 1.3-2 em long.. el posterior apenas
emarginado. el anterior profundamente trilobado.
Еч
con los lóbulos de 11 mm long. por 4—5 mm lat.
Estambres con filamentos de 8 mm long. insertos en
el ápice del tubo, v anteras con las tecas a distinta
altura: la superior de 2 mm long., la inferior apenas
más larga. v apendiculada en la base. Cápsula obo-
voide, de 1.5 em long. por 0.6 em diám., sólida y
lateralmente estrechada en la mitad inferior, inflada
y 4-seminada en el ápice. Semillas subesféricas, de
lisas y lustrosas: jaculato-
—3 mm diám., castañas,
res de 2.5 mm long., obtusos.
Figura 8.
Ilustraciones.
hábitat Sudeste de
Paraguay y norte de la Argentina. en
Distribución, y fenologta.
Bolivia, a
región del Chaco oriental. Habita en bordes y abras
de bosque chaqueño, en lugares soleados con inun-
dación periódica, y florece en primavera, verano y
otoño, de octubre a mayo.
Justicia dumetorum se caracteriza por las flores
rojas o moradas, de corola larga (3—4 cm). aparen-
temente polinizadas por picaflores. Existen algunos
ejemplares de Paraguay y Bolivia que se citan con
asterisco (*) en la lista de material (Fiebrig 5074.
Fiebrig 5.138, Beck 5373 y Troll 1688) que se di-
ferencian solamente por las corolas menores (de ca
2 em long.), lilacinas y de tubo basal más corto,
con aspecto de ser polinizadas por abejas о abe-
jorros (melitófilas). Estos ejemplares se asemejan
bastante a Justicia furcata Jacq., especie descripta
para México, y en algunos casos fueron determi-
n ejemplares
~
nados con ese nombre por Lindau
de herbario (leg. Fiebrig 5074 y s.138).
Justicia
furcata no habita en América del Sur (Wasshausen,
com. pers.). y además estos ejemplares se diferen-
cian de Justicia furcata del sur de Nortemérica у
de América Central por las brácteas generalmente
más anchas v las inflorescencias con menos flores.
Sin embargo estos ejemplares coinciden mucho con
especies afines. descriptas para. América del Sur
como Justicia peruviana Lam. (ilustrada en Cava-
nilles, Icon. 1: 17, tab. 28. 1791, y Curtis Botanical
Magazine 11—12: 430.
y Justicia lithospermifolia Jacq.
1797), Justicia carthagine-
nesis Jacq..
Todas estas especies afines son muy variables, y
en ellas se han encontrado formas con flores de
distinto tamaño y diferente colorido (Leonard.
1951-1958: Daniel, 1995).
con Daniel (1995) en que el conjunto de estas es-
Debido a ello coincido
pecies de Justicia sect. Simonisia indudablemente
aliadas a Justica carthaginensis Jacq. 1 Justicia
furcata Jacq. deberían ser objeto de un estudio de
material abundante de toda su área de distribución,
que incluya un análisis detallado de la forma y el
color de la corola además de la distribución y el
ambiente habitan.
que requisitos. indispensable
para intentar diferenciarlas. Hasta tanto se realice
248
Annals of the
Missouri Botanical Garden
— 8. Justicia dumetorum. —A. Rama con flores.
|
А OF )
B. Flor con bráctea, bractéolas, cáliz, corola, estambres y
estilo. —C. Corola abierta con androceo. —D. Estilo y estigma. —E. Cápsula abierta. —F. Semilla. Schulz 8673 (SI)
Saravia Toledo 10779 (SI).
ese estudio, trato al material de flores cortas y li-
lacinas que aparece en Bolivia y Paraguay como
parte de la variabilidad intraespecífica de Justicia
dumetorum.
Se aclara que el supuesto isotipo del nombre
Justicia kerri, Kerr 108, que se encuentra en el her-
bario de Edimburgo (E). pertenece a Justicia bra-
siliana Roth y no concuerda con la descripcion
original de Justicia kerri. Probablemente esté erró-
neamente etiquetado.
Volume 89, Number 2
2002
zcurra 249
Justicia en Sudamérica Austral
ARGENTINA. Corrientes:
rs Bernardello y Galetto 800 (SI); Dpto. San
losme, 5 km E de Paso de la Patria, Krapovickas y Cris-
E 11889 (SI). Chaco: — Resistencia, —
ras, Meyer 16235 (LIL); Dpto. 1° de mayo, С‹ 1 Be-
nítez. Schultz 8673 B. i r
monte, Jörgensen 2332bts (SD):
Clorinda a Formosa, Km 3, Morel 17
Tigre, Parodi 8511 (LIL). Salta: Rivadavia, 18 k
de Dragones, Saravia Toledo 10779 (51); Dpto. Orán, Em-
barcación, km 1317, Schreiter 10866 (SI). BOLIVIA.
Beni: Ballivián, ерйп, еп la zona de influencia del Rio
Yac ‘uma, и húmeda, Beck 5373 * (flor — и
ep.: Cañamina, Río La Paz, Troll 1688* (M).
Boquerón: Colonia Menno, ect Are-
nas 1097 (LIL), Arenas 1457 (US); Dr. ro P. Peña, 10
km al W del pueblo, Spichiger et al. 2217 (G, US); Fila-
delfia, Hahn 787 (US). Chaco: Mayor Pedr
Material seleccionado.
Casa Cué,
c
ro Lagere nza.
kn
Salado, de Limpio a Embose ada, Arbo
(CTES): Рае Correo, montes húmedos, orilla Río Salado.
Hassler 1650 (SI). Concepción: Zw. Rio 2 und Rio
Aquidaban, V dla Sana, Fiebrig 5074* (G, К
Presidente Hay
Laguna Yaraguf, 7 km «
. dep.: s.
loc., Fiebrig s.138* (G).
(BM); Colonia Menno, Я
Arenas 143 (US); in regione cursus inferioris fluminis
'omayo, M 81 (G, MO); pozo € — eds les y
gen 5563 (FCO): Villa Hayes, Zardini et al. 2592 (М‹ у).
Zardini et p 2593 (FCQ, MO, US
11. Justicia floribunda (C. Koch) Wasshausen.
Darwiniana 35: 151. 1998. Libonia floribunda
C. Koch, Wochenschr. Ver. Bef. Gartenb. Kon.
Preuss. Staat. 6: 266. 1836. TIPO: ilustración
del ejemplar cultivado en el Jardín Botánico
м
de Bruselas sobre el que se basó la descrip-
ción original en Morren, Belg. Hort. 14: 12.
1864 (lectotipo, — por Wasshausen en
Peixoto et al., 1
Seric sige i ig Nees, in Martius, Fl. Bras. 9: 109.
1947. Justicia — (Nees) Griseb., Abh. Ko-
i Wiss. Gottingen 24: 262. 1879, non Justicia
а Vahl, Eclog. Am. 1: 2. 1796. Jacobinia
pauciflora (Nees) Lindau. in Engler & Tus Nat.
flanzenfam. asl. 1895.
Wasshausen, Baileya 19: 3.
loc., Sellow s.n. iim
K!).
—
—
Justicia. rizzinii
1973. TIPO: Brasil. 5.
fot. F 89147; prob-
В destr.,
p.
ables isosintipos, El,
Sericographis pauciflora — a — Nees, nr
tius, Fl. Bras. 9: 109. TIPO: Brasil. S. loc.,
Tweedie s.n. (sintipo, oo
Arbusto ramoso de 0.50—1(—2) m alt.,
cilíndricas. pubérulas. Hojas con pecíolos de 0.4—
con ramas
1 em long. y lámina ovada, elíptica a obovada, muy
variable en tamaño, de 1.5-7 cm long. por 0.8-2.
mm lat., las de cada par desiguales, las superiores
mayores, obtusas у cuneadas, enteras a levemente
aserradas, discolores, glabras o leve y cortamente
pubescentes. Espigas paucifloras frecuentemente
reducidas a flores solitarias sobre pedtinculos dé-
biles hasta de 3 em long. en las axilas de las hojas
superiores, con brácteas y bractéolas triangular-
anceoladas, ca. de 1 mm long. por 0.5 mm lat.,
glabras a cortamente ciliadas. Cáliz 5-partido de 5
mm long., segmentos lanceolados, de 3.5 mm long.,
acuminados, pubérulos. Corola de 2-3 cm long.,
roja y con la mitad al tercio superior amarillo, tubo
~
‘a. de 20 mm long. por 5 mm diam. en la fauce,
por fuera pubérulo, con algunos pelos glandulares,
interiormente glabro; los labios oblongos, ca. de 5
mm long., el posterior erecto y obtuso o emargi-
nado, el anterior erecto o algo patente y trilobado,
con los lóbulos de ca. 2 mm long., triangulares.
Estambres inclusos debajo del labio superior, fila-
mentos pilosos en la parte media, anteras ca. de 2
mm long.. algo superpuestas, oblicuas, caudadas.
Ovario glabro o glanduloso-pubérulo en la porción
apical. estilo glabro. Cápsula claviforme de 12-17
mm long. por 4 mm lat. y 2 mm de espesor, glabra
o glanduloso-pubérula; semillas aplanadas, subor-
biculares. ca. de 2 mm diám., castañas, levemente
rugosas: retináculos, ca. de 2 mm long., obtusos.
Ilustraciones. Peixoto et al., 1998: 153.
Distribución, hábitat y fenología. Sur de Brasil
v noreste de Argentina (este de Misiones). Tal vez
también se encuentre en Paraguay oriental. Habita
en selvas y bosques húmedos, especialmente sobre
suelos periódicamente inundados, entre el nivel del
mar y 1400 m de altitud. Florece en invierno, de
mayo a agosto, y sus flores rojas de ápice amarillo
probablemente sean polinizadas por picaflores.
Usos. Esta especie se cultiva como ornamental
en regiones cálidas del mundo por sus llamativas
rojas (Hay & Beckett,
pauciflora). Las flores de esta especie se producen
flores 1975. sub Jacobinia
principalmente en invierno, y las plantas se repro-
ducen fácilmente por gajos no floríferos y semillas
(Parodi & Dimitri, 1980, sub Jacobinia pauciflora).
Vombre vulgar. “Bandera española” en Argen-
tina (Parodi & Dimitri, 1980, sub Jacobinia pau-
ciflora).
Justicia pauciflora se caracteriza por las flores
ornitófilas con la corola roja de ápice amarillo. Pa-
rodi y Dimitri (1980, sub Jacobinia pauciflora) ci-
taron a esta especie frecuentemente cultivada en
\rgentina como originaria de Brasil, por lo que
cabe la posibilidad de que las poblaciones silves-
tres de Misiones hayan resultado de escapes de
cultivo. Esto ha sucedido con otras especies de
Acantáceas neotropicales cultivadas, que se han
asilvestrado en regiones alejadas de su área de dis-
tribución original, como por ej. Justicia. carnea
Lindl. (en Nueva Zelandia. Webb, 1995) y Ruellia
breviflora (Pohl) Ezcurra (en Asia y Australia, Ez-
250
Annals of the
Missouri Botanical Garden
curra, 1989). Sin embargo, es probable que las po-
blaciones silvestres de Justicia floribunda del este
de Misiones se encuentren dentro del área de dis-
tribución natural de la especie, ya que ésta fue
descripta originalmente para regiones contiguas del
sur de Brasil (Nees, 1847a, 1847b; Wasshausen &
Smith, 1969) y
el siglo XIX, como lo demuestra el ejemplar colec-
ademas ya existfa en Misiones en
cionado por Niederlein en 1887.
Material seleccionado. ARGENTINA. Misiones: Ber-
nardo de lri rigoyen, Hunziker, J. 950, 959 (LIL); Varana,
territorio de Misiones, Niederlein 1776 (LIL); Guaraní.
Predio Guaraní, Arroyo Soberbio, borde del arroyo en bar-
ranca inundable, — et al. 5647 (SD; Gral. Manuel
Belgrano, de Bernardo de Irigoyen a San Antonio pasando
Salto Andresito, Zuloaga 2111 (SD.
12. Justicia gilliesii (Nees) Bentham, in Bentham
« Hooker, Gen. Pl. 2(2): 1109. 1876. Adha-
in DC., Prodr. 11: 395.
. TIPO: Argentina. San Luis: Gillies s.n.
toda gilliesii Nees,
1847
(sintipo, K!; isosintipo, E!); Santiago del Es-
Tweedie 1260 p.p.,
mezcla con Justicia tweediana (sintipo, K!).
tero: "St. lago de Cordova,"
Justicia е Hieron., Bol. Acad. Nac Córdoba
4(1): 61. 1881. TIPO: Argentina. oh duos De}
zo. — cia Maradona, Ene. 1876, Echegaray
n. (holotipo, CORD!)
Sufrátice ramoso de hasta 0.60 m alt., con ramas
ascendentes o rastreras, cilíndricas, estriadas, pu-
bescentes. Hojas pecioladas con pecíolos de 0.5-2
cm long., ovadas a elípticas, de 2.5-3 em long. por
1.8-2.5 em lat.. agudas и obtusas y apiculadas, cu-
neadas en la base, membranáceas, generalmente
pubescentes. Flores sésiles o cortamente pedicela-
das, con dos bractéolas en la base, solitarias en las
axilas de brácteas, formando espigas de 4-12 cm
long.; bractéolas elípticas de cerca de 1 cm long.
por 2 mm lat.; brácteas de 9-12 mm long. por 4—
5 mm lat., ambas glanduloso-pubescentes, ciliadas
en el margen. Cáliz profundamente 5-partido, con
los lóbulos angostamente ovados a lanceolados, su-
biguales, de 4 mm long.. foliáceos, generalmente
glabros. Corola celeste-lilacina a blanca de 2 cm
long., pubérula, con tubo corto de cerca de 1 ст
long., ensanchado en una garganta de aproxima-
damente 0.8 cm diám., los labios de 1—1.3 cm
long., el anterior trilobado hasta la mitad, con la
fauce blanca, convexa y transversalmente rugoso-
venosa, con puntos morados. Estambres insertos en
la base de la garganta, las anteras con las tecas
subparalelas, una por encima de la otra, la inferior
generalmente apendiculada en la base. Cápsulas
obovoides, pubescentes, de 1.2-1.5 em long. por
5-7 mm ancho, con la mitad inferior sólida y la-
teralmente comprimida. Semillas generalmente 4,
suborbiculares, aplanadas, de 3—4 mm diám., par-
das, diminutamente rugoso-tuberculadas; retinácu-
los obtusos.
Ilustraciones. Figura 9.
Noroeste vy
Distribución, hábitat y fenología.
centro de la Argentina, desde Catamarca, Tucumán,
La Rioja, San Juan y Mendoza hasta Córdoba, San
Luis y Santiago del Estero. Se la encuentra en mon-
tes áridos entre 300 y 1000 m s.m. Florece en ve-
rano y otoño, de diciembre a abril.
Nombre vulgar. “Albahaca de vaca” (Burkart,
1943, sub Justicia echegarayi).
Justicia gilliesii ha sido citada como de impor-
tancia como planta forrajera (Ariza-Espinar, 1971),
al igual que varias otras Acantáceas del sur de Su-
damérica (Burkart, 1943).
Material seleccionado. ARGENTINA. Catamarca:
La Paz, el Divisadero, 13 Abr. 1947, Brizuela 1132 (LIL);
Ancasti, El Cajón, Brizuela 1266 (LIL); Ambato, entre La
Puerta y Huaicama, Cristóbal 442 (LIL); Capital, Cata-
marca, Nicora s.n. (SI 18656); Camino a El Juncal, Soria-
no 628 (SI); Paclín, entre Catamarca y Cuesta del Totoral,
Villa — 1119 (LIL). Córdoba: Tulumba, Salinas
Grandes . V. Mansilla, Hunziker, А. 10985 (CORD);
Cruz del Eje. — Cruz del Eje y Villa S
mbre, Nicora 1793 (SI).
oja: San Martín, Bajo s Campo Balde El Tala,
uso 17, — 4044 (CORD); Chamical, Sierra de
are C — ily la ipu Biurrun
1372 (CORD); Capital La Rioja, Burkart 12531 (SI); en
las cercanías de arlos, camino a Córdoba, Hierony-
mus y Niederlein p p ORD); Sierra de Velasco, —
al Dique Los Sauces, Hunziker, А. 4721 a IL); A. V.
a 29, al S de Punta de los Llanos, * el
desvío a Tama у Carrizal, — А. 15096 (CORD);
Famatina, Guanchin, Venturi 6 (SI). Mendoza:
Paz, Desaguadero, Ruiz Leal Te (LIL). San Juan: Va-
lle Fértil, N de Usno, Kiesling 8961 (SI); Valle Fértil,
camino a La Majadita, Kiesling 5012 (SI). San Luis: Cap-
ital, San. Luis, alrededores, Troncoso s.n. Pe — Bel-
grano, Alto Pencoso, Bruch y Carette rm
Sierra de Las Quijadas, Qda. del Alambre, Fi W de
los Llanos, falda F
&
» 2.
I
—
n Pierotti s.r
mán: Vipos, Lillo 7894 (L n ); Dpto. — ‘as, Tapia, Ven-
1814 (SI).
—
—
—
y
A
oe
—
~
D
-
—
=
ae
^
D
DP
~
=
n
z
~
=
-
=
~
Y
—
=
turi
13. Justicia glaziovii Lindau, Bull. Herb. Boiss.
483. 1895. TIPO: Brasil. Glaziou 13073
(lectotipo, designado por Graham (1988), K!).
Rhytiglossa — Nees, in Martius, Fl. Bras. 9: 129.
7, non Justicia paniculata Burm. f. 1768. TIPO:
Brasil. 8 ahia: Ilheos, Blanchet 2979 (isotipo, BM!).
Lophothecium paniculatum Rizzini, An. 5 Bot. Hio de
aneiro 8: 336, t. 5. 1948, ' “Minas ( ›е-
rais: Ituitaba, Macedo 1123 (isotipo, BM!)
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
1 mm
Е
мдү у
М Y `
C
Figura 9. Justicia gilliesii —A. Planta con flores; —B. Flor con bráctea. bractéolas. cáliz. corola, estambres v
estilo. —C. C ў Cá )
Márquez s.n. (SI 45519)
Justicia velascana Lindau. Bull. Herb. Boiss. 3: 184
1895. TIPO: Bolivia. “Prov.
m, Ago., Kuntze s.n. (holotipo, B destr.. fot
isolipo, NY!).
‹
elasco orientali", 200
p К 0070|.
. OOO...
Arbusto ramoso con tallos ramificados, cilíndri-
cos, сопатепіе pubérulo-tomentosos. Hojas con
pecíolo de 0.5-1.5 em long., y lámina ovada. de 5—
8(-15) em long. у 2.54.5 em lat. (la de las hojas
apicales menor). aguda en el ápice, cuneada a re-
dondeada en la base v algo decurrente. sobre. el
pecíolo, glabra o pubérula en el haz. pubérula o
cortamente pubescente en el envés, en especial so-
Corola abierta con androceo. —D. Estilo y estigma. Е. Cápsula abierta. —F. Semilla. Kiesling 8961 (S1).
bre las venas. Inflorescencias paniculiformes en las
axilas de las hojas superiores, amplias, tenues. de
espigas laxas reunidas en racimos compuestos ge-
neralmente más largos que las hojas, con los raquis
pubérulos: brácteas anisofloras. lineales o triangu-
ares, diminutas. de 1-2 mm long. y 0.5 mm lat.,
elabras o pubérulas: bractéolas similares, más an-
gostas. Cáliz profundamente 5-partido, de ca. 4 mm
long., con los segmentos lineales, de 3-4 mm long.
por 0.5 mm lat., glanduloso-pubérulos. Corola color
crema de 0.8-1.3 em long., con tubo acampanado
de 5-8 mm long. y 3-5 mm Фат. en la fauce. los
Annals of the
Missouri Botanical Garden
labios de ca. 5 mm long. el posterior ovado, de 4—
5 mm lat., biapiculado, el anterior trilobado, con
los lóbulos de ca. 2 mm long. por 2.5 mm lat. Es-
tambres exertos hasta el ápice del lóbulo posterior,
con las tecas superpuestas, oblicuas, de menos de
| mm, separadas por el conectivo, la inferior más
pequefia y con un apéndice glandular claviforme
en la base. Cápsulas angostamente claviformes de
ca. | cm long. y 2 mm lat.,
sólido y lateralmente comprimido y la porción su-
perior engrosada, algo estrechada en su parte me-
dia, pubérulas. Semillas de 1.5 mm diám., tuber-
culado-verrugosas; retináculos de ca. 1 mm long.
Ilustraciones. — Wizzini, 1948: 339 (sub Lopho-
tecium paniculatum).
Sur de Brasil,
este de Bolivia, y noreste del Paraguay. Habita en
Distribución, hábitat y fenología.
bordes de bosque. Florece en verano y otoño, prin-
cipalmente de enero a marzo.
Justicia glaziovii se caracteriza por las inflores-
cencias paniculiformes amplias y tenues, con flores
color crema de cerca de 1 cm long., y es una es-
pecie relativamente rara que ha sido poco colec-
cionada.
Material seleccionado. PARAGUAY. Amambay: in
altiplanitie et declivibus "Sierra de Amambay”, Rojas
10864 (BM, G, K, M, NY, P, W); Sierra de Amambay, in
via sylvatic a, Picada Yatobó, Cerro Рога, Rojas 104
; Zw. Río Apa und Río Aquidabán, Ir а,
Fiebrig 4811 (BM, G, K). Concepción: Estancia Potre-
rito, Basualdo 3485a (FCQ).
14. Justicia glutinosa (Bremekamp) V. A. W.
Kew Bull. 43: 613. 1988. Sarotheca
glutinosa Bremek., Proc. Kon. Ned. Akad.
Wet. C, 72: 426. 1969. TIPO: Bolivia. Brooke
5677 (holotipo, BM no visto, fot. US!).
Graham,
Sarotheca elegans Nees, in Martius, Fl. Bras. 9: 113.
‚ non Justict ia — Bêauv. 1810. Ju sa-
— . W.
— taham, gs — 43: 614.
TIPO: brasil: Goiás . Felis E E =
ciras, Pa 1989 holotipo, №, P F 327
Hierba sufruticosa de 0.5-1.5 m alt.. con tallos
subcilíndricos, largamente hirsuto-pubescentes
cuando jóvenes, glabrescentes a la madurez. Hojas
con pecíolo de 1—4 cm long., y lámina ovada a
anchamente elíptica de 7-15 em long. por 3-5 cm
lat., aguda y acuminada en el ápice, cuneada y
decurrente en la base, pubérula en el haz y más o
menos hirsuto-pubescente en el envés, especial-
mente sobre las venas principales. Inflorescencias
en espigas densas reunidas en racimos peduncu-
lados en las axilas de las hojas superiores: pedún-
culos y raquis glanduloso-pubescentes; brácteas
con el tercio inferior
opuestas isofloras o alternadamente anisofloras,
anceoladas, de 2.5-3 mm long. y | mm lat., glan-
duloso-pubescentes; bractéolas lineales, algo me-
nores. Cáliz profundamente 4-partido, con los seg-
mentos lanceolados de 4—6 mm long. y 1.5 mm lat.,
densamente glanduloso-pubescentes. Corola de ca.
| cm long., blanca con marcas lilas en la fauce,
externamente glanduloso-pubescente, el tubo le-
vemente obcónico, de 4-6 mm long., y labios de
4-5 mm long., el posterior ovado, de 2 mm lat.,
agudo, el anterior trilobado, con los lóbulos de 1.5—
2.5 mm long. redondeados. Estambres exertos hasta
el ápice del labio posterior, las anteras con la teca
superior oblicua, de ca. 1 mm long., la inferior algo
menor y apendiculada en la base. Cápsula clavi-
forme de 1.5-2 cm long. y 0.4 cm diám., densa- y
cortamente glanduloso-pubérula. Semillas marca-
damente aplanadas, de 3 mm diám., papilosas; re-
tináculos de menos de 2 mm long.
Ilustraciones. Nees, 1847a: tab. 18 (sub Saro-
تس
theca elegans).
Distribución, hábitat y fenología. Perû, Bolivia,
extremo norte de Argentina, sudoeste de Brasil y
nordeste de Paraguay. En Paraguay habita en bor-
des de selvas riberefias rodeadas de campos cerra-
dos. y florece principalmente en primavera, de
agosto a noviembre.
El nombre Justicia sarotheca se reduce a la si-
nonimia de Justicia glutinosa por primera vez. Esta
especie se caracteriza por sus hojas grandes y flores
medianas, de cerca de 1 em long., en inflorescen-
cias glandulosas de espigas densas reunidas en ra-
Existe una ilustración de esta es-
Flora
cimos axilares.
pecie (como Sarotheca elegans) en
Brasiliensis (Nees, 1847a).
Material seleccionado. ARGENTINA. Salta: Dep.
San Pu Quebrada Astillero, Schulz 5484 (CTES).
PARA GUAY. Amambay: Cerro Corá, en arroyo interrum-
pido en selva, Schinini y Bordas 20344 (CTES, G, US);
Parque Nacional Cerro Cora, bosque húmedo de hasta 25
m de altura al lado = Arroyo Aquidaban-Ningiif, suelo
Brunner 1442 (G, MO): Parque Nacional Cerro
sora, cerca de la casa forestal, en una maleza algo
meda al lado de | la carretera, Ferndndez Casas 4114
rque Nacional Cerro Cora, ape: 5 del rend
Aduidaban- Aguí, Soria 1727 (FCQ, G, MO); Parque Na-
ional Cerro Сога, selva marginal i — Aquidabán-
N iguí, Ferrucci, Vanni y Ferraro 693 (CTES); Parque Na-
cional Cerro Cora, Mereles 2054 (G).
arenoso,
Graham, Kew
1988. s — ylax umbrosus
Prodr. 313. 1847, non Jus-
1841.
TIPO: Argentina. Tucumán: San Javier, Twee-
die 1262 (lectotipo, aquí designado, К!).
15. Justicia goudotii V. A. W.
Bull. 43: 603.
Nees, in DC.,
ticia umbrosa a 2: Hare: 79.
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
pcr boliviensis Lindau, Bull. Herb. Boiss. 3: 492.
non Justicia boliviensis (Bremek.) V. А. W.
1988. b IPO: Bolivia.
i. Kew Bull. 43: 613.
anta Cruz: 1000 m, Mayo 1892, Kuntze s.n. (sintipo,
p d fot. Е 88971; isosintipo. NY!): Brasil. Pa-
í, s. col. 466 (sintipo, B. destr.
lla vestitus Rizzini, Dusenia 3(3): 191. 1952,
non Justicia — Schult., — 1: 145. 1822.
TIPO: Brasil. Paraná: Parque Nacional do Iguazú. 3
Abr. 1949, Duarte y Pereira 1927 — RB no
visto).
Hierba perenne, rizomatosa en la base, con tallos
erectos subcilíndricos, a veces decumbentes y ge-
niculados, de hasta 80 cm long..
opuestas con pecíolo de 3-10 mm long. y lámina
2-9 cm long. por 1.5—3.5
pubérulos. Hojas
ovada de 2 cm lat., acu-
minada en el ápice, cuneada y algo decurrente en
a base, entera, pubérula en el haz, pubescente so-
bre la nervadura en el envés. Inflorescencias for-
madas por flores sésiles con dos bractéolas en su
base, dispuestas en el eje de la inflorescencia for-
mando una espiga unilateral. Espigas terminales o
en las axilas de las hojas superiores, a veces en-
riquecidas por espigas derivadas de yemas super-
numerarias. Brácteas y bractéolas lanceoladas, rí-
gidas, de 4-6 mm long., ciliadas, con el nervio
ее
medio prominente en el dorso. Cáliz de cerca de €
mm long., con los 4 segmentos rígidos. lineales,
subulados, dorsalmente pubescentes sobre el ner-
vio medio, ciliados. Corola pálido lilacina a violá-
cea de 1.5-2 cm long., con tubo muy angosto de
1-1.5 em long., estrecho en la garganta, el labio
superior erecto, de 4 mm long. por | mm lat., el
labio inferior ancho, trilobado, de cerca de 5 mm
long.. con estrías blancas en la fauce. Estambres
con filamentos de aprox. 5 mm long.: teca apical
perfecta, de contorno ovado, de 1 mm long., la otra
inserta más abajo en el filamento, semiestéril, más
angosta y basalmente caudada. Cápsula de contor-
no obovado, de aprox. 7 mm long. por 2.5 mm lat.,
cortamente sólido-estipitada en la base. glabrescen-
te y de paredes finas. Semillas 4, aplanadas. de
mm diám.. con pelos cortos y gloquidiados: reti-
náculos de 1 mm long., obtusos.
Ilustraciones. Ezcurra, 1993a: 344.
Distribución, hábitat y fenología. Colombia.
Bolivia, Brasil, Paraguay y norte de la /
Habita en el interior o en los bordes de selvas me-
Arge ntina.
sófilas y bosques xerófilos de tipo Chaqueño, fre-
cuentemente en ambientes alterados de planicies y
serranías hasta 1300 m aprox. Florece a fines de
verano otoño e invierno, de enero a julio.
“Uchu yuyo” en el noroeste de
1943, sub Chaetothylax um-
Nombre vulgar.
Argentina (Burkart,
brosus).
Esta especie se describió bajo el nombre de
Chaetothylax umbrosus Nees sobre la base de dos
sintipos, uno de Colombia (Goudot s.n.. K!) y otro
de Argentina (Tweedie 1262, K!). Graham le puso
un nombre nuevo al pasarla a Justicia, Justicia gou-
dotii Graham, por ya existir el nombre anterior Jus-
ticia umbrosa Benth., pero no designó un lectotipo.
Debido a que los sintipos pertenecen a dos especies
diferentes, una de Colombia y otra de la región
subtropical de Sudamérica (norte de Argentina, su-
deste de Bolivia, Paraguay y sudoeste de Brasil):
1) Selecciono el ejemplar de Tweedie de Argentina
como lectotipo de Chaetothylax umbrosus (= Jus-
ticia goudotii). La elección de este ejemplar se basa
en que, ante la falta de otras evidencias en el pro-
tólogo. el epíteto umbrosus que eligió Nees para la
especie coincide con los datos del ejemplar de
Tweedie “muy abundante en los bosques mas som-
bríos de San Javier” (traducido). Por otro lado, la
elección de este ejemplar preserva el uso actual del
nombre, ya que la especie de la region subtropical
de Sudamérica ha sido identificada con los nombres
Chaetothylax umbrosus y Justicia goudotii en varias
publicaciones (por ej. Ezcurra, 1993a, 1999), y está
identificada con estos nombres en numerosos ejem-
plares de herbario. (2) El ejemplar de Goudot de
Colombia pertenece a la especie determinada por
Leonard (1951-1958) como Chaetothylax huilensis,
que se diferencia de Justicia goudotii por las inflo-
bractéolas y
rescencias más anchas, las brácteas,
lóbulos del cáliz más largos, y las cápsulas mayo-
res.
Material seleccionado. ARGENTINA. Catamarca:
del Alto. Balcagua, Venturi 7080 (SI). Corrie — Ca-
mino a Sta. Cat: Soriano 1665 (SI). Chaco: Las Bre-
ñas, legname y — 8704 (LIL): * е PN Cha
Маша 927 (SI). Formosa: Km 39 de j se
Lagun a rs Meyer
Venturi 5301 (LID):
dina,
Calile gua,
da, Legname 6857 (LIL); Capita Ruta 51, km
20, = km al E de Salta, Novara 3502 (SI); Anta, Tacuí,
15 Km al S de Macapillo, límite con Santiago, Novara y
Terne 10284 (LIL): Orán, San Ignacio, Pierotti 83 (LIL);
Rosario de la Frontera, pon 6184 (LIL). Santa Fe:
Villa Guillermina, Meyer 6 (L IL. Santiago del Es-
ero: Choya, La Punta, E Del Mogote, Qda. De Cala-
pachin, Frenguelli 128 (SI); Villa La Punta, Sa. de Gua-
saván. Ulibarri 1509 (SI). T та Rio Chico, Qda. de
Marapa, — — s.n. (LH í, Parque Aconquija, de
la Sota 1138 (LIL); Tucumán, 5 s.n. (К); Famaillá,
La died "Ree 1724 . PARAG = Guairá:
Cordillera de Ybytyruzú, Cerro My My, camino a Cerro
Polilla, Zardini y Velásquez 13803 (MO). Р Жыр a
— 1 Chaco, Santa Elisa, Rojas 2818 „СЕ,
LIL, MO, NY, T in — cursus inferioris ША. Pil-
'omayo, Us jas 155 (G); 50. Soria 280 (G); Route
— :haco, pk E Billiet y y Jadin 3057 (BM, G NY).
їл
=
254
Annals of the
Missouri Botanical Garden
16. Justicia hassleri (Lindau) V. A. W. Graham.
Kew Bull. 43: 615. 1988. Beloperone hassleri
Lindau, Bull. Herb. (App. 1): 30.
1898. TIPO: Paraguay. "In дитей pr. Hacu-
rubi,” Hassler 1089 (holotipo, С: isotipos.
MO!, NY!, P!).
Boiss.
e marginata Lindau, Bull. Herb. Boiss. 3:
19 895, non Justicia marginata Lindau, 1894.
puede rb cad A. W. Graham, Kew Bull.
3: O15 TIPO: Paraguay meridional, Se dm
1892, — ге s.n. кош B destr., fot. F 8%
NY".
isotipos, С!,
—
Hierba sufruticosa, rizomatosa en la base, con
tallos erectos subcilíndricos, de hasta 40 em long.,
generalmente glabros. Hojas sésiles, opuestas, con
Amina ovada de 3.5-6(-9) ст long. por 1.3-3 cm
lat., entera,
rígida, coriácea, aguda y acuminada en
el ápice, redondeada a cordada en la base, con el
margen escarioso y las venas prominentes en el en-
vés, glabra. Flores sésiles dispuestas en espigas
denm sésiles, terminales o en las axilas de las
hojas superiores, con brácteas anisofloras. Brácteas
y bractéolas angostamente lanceoladas a lineales,
imbricadas, rigidas, de 8-15 mm long. por 1.5-2.5
mm lat., largamente acuminadas, escariosas, con el
margen blanquecino. Cáliz 4-partido de 10-15 mm
long., con los 4 segmentos rígidos. lineales, subu-
lados, albo-marginados. Corola pálido lilacina a
violácea de ca. 2.5 em long., con tubo muy angosto
de 1.5-2 em long., estrecho en la garganta, el labio
superior erecto, de 6 mm long. por 1.5 mm lat., el
labio inferior ancho, trilobado, de cerca de 8 mm
long., con los lóbulos de 5 mm long. por 4 mm lat..
el medio algo mayor. Estambres con filamentos de
aprox. З mm long.: teca apical perfecta, de contorno
ovado, de 1 mm long., la otra inserta más abajo en
el filamento, más angosta y basalmente caudada.
Cápsula de contorno elíptico, de 6 mm long. y 5
mm grosor, cortamente estipitada en la base, glabra
y de paredes finas. Semillas 4, aplanadas, de 1 mm
diám., corta y densamente papiloso-pilosas, pardo-
oscuras; retináculos agudos.
Ilustraciones. Figura 10.
Distribución, hábitat y fenología. Sur de Brasil
(Mato Grosso do Sul) y Paraguay central. Habita en
bordes de bosques. Florece a lo largo del año, prin-
cipalmente en invierno, de j junio a septiembre.
Justicia hassleri se asemeja a J. alopecuroidea T.
F. Daniel (Daniel, 1990), especie de México, lo que
sugiere relaciones filogenéticas estrechas entre al-
gunas especies de Justicia de Sud y Norteamérica.
las que deberían estudiarse.
Material PARAGUAY.
seleccionado. Cordillera: 3
km desvío ruta 2 a Piribebuy, Ferrucci 758 (CTES); in
— sto Hacurubi, Hassler 3053 (G, K, NY, P, №); in du-
eto pr. Itacurubi, Hassler 1138 (G); San Bernardino, ori-
llas montes, Hassler 138 (SI). Paraguarí: Cerrito, in du-
metis, Osten y Rojas 8978 (С): Cerro Hu, pres de
Paraguarí, Balansa 2456 (С. К); Cerro Mbatoví. camino
al arroyo Mbatoví, Basualdo 2658 (FCQ); Cerro —
forest along N slope, Zardini y — 9943 (MO); Cerro
Paraguarí, Hassler 916 (BM, W), Fiebrig 916 (E, G, К,
); in re gione collium Cerros de Раа, Hassler 6476
(BM, G. K, LIL, NY, W). S. dep.: Paraguaria Centralis,
Hassle 3057 (BM); Piura Centralis, reg. lacus Ypa-
caray, Hassler 12530 (BM, E, G, LIL, MO, NY, US); Sed-
paraguay, Auntze s.n. (G)
Justicia hunzikeri Ariza Espinar, Kurtziana
6: 92. 1971. TIPO: Argentina. La Rioja: San
Martín, Sierra de Ulapes, falda E, frente a Ula-
pes, 25 Mar. 1958, Hunziker, A. y Caro 13560
(holotipo, CORD!).
Arbusto de 0.50—1.50 m alt., con ramas erectas,
cilíndricas, las jóvenes amarillentas y cuadrisul-
cadas, con estrías marcadas entre las costillas, ge-
neralmente glabras. Hojas cortamente pecioladas,
angostamente ovadas a elípticas, de 2—4 cm long.
por 0.3-1.5 cm lat., agudas, cuneadas en la base,
generalmente glabras, pubérulas solamente sobre la
vena media en el envés. Flores sésiles o cortamente
pediceladas en espigas terminales y axilares de 3—
5 cm. long.; brácteas ovadas o e otic as de 1.4 cm
em lat., bractéolas lanceoladas de cer-
ca de 1 em long. por 0.2 cm lat.
long. por 0.7
Cáliz profunda-
mente 5-partido, con los lóbulos angostamente ova-
dos a lanceolados, subiguales, de 3—4 mm long..
foliáceos, generalmente glabros. Corola lilacina de
2-2.5 em long., pubérula, con tubo corto de cerca
de | cm long., ensanchado en una garganta de
aproximadamente 0.8 ст diám.,
1.3 em I el anterior trilobado hasta la mitad,
con la fauce blanca,
los labios de l-
ong.,
convexa y transversalmente
rugoso-venosa. Estambres insertos en la base de la
garganta, las anteras con las tecas subparalelas,
una por encima de la otra, la inferior generalmente
apendiculada en la base. Cápsulas obovoides, ro-
bustas, glabras, de 0.8-1 cm long. por 5-6 mm
diám.. con la mitad inferior sólida y lateralmente
‚ suborbicu-
diminutamente rugoso-tuberculadas:
comprimida. Semillas generalmente 4
ares, retiná-
culos obtusos.
Ilustraciones. — Ariza-Espinar, 1971: 93.
Endémica de
Argentina, se la encuentra en La Rioja, San Juan
Distribución, hábitat y fenologia.
Habita frecuente-
mente en matorrales serranos sobre suelo rocoso a
aprox. 600—800 m s.m. y florece en primavera v
verano, de noviembre a marzo.
y San Luis, en regiones áridas.
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
и.
tilo. 1. Cápsula abierta. —D. Semilla. Hassler 156 S
Justicia hassleri. — А. Rama con flores.
Justicia hunzikeri se asemeja a Justicia tweedia-
na en las corolas lilacinas, pero se diferencia por
las flores en espigas secundifloras y las semillas
rugoso-tuberculadas sin engrosamiento en el mar-
gen, en lo que se asemeja a Justicia xylosteoides.
Justicia hunzikeri habita un área restringida al ex-
tremo NW de 5
Rioja y posee distribución simpátrica con J. xylos-
San Luis. E de San Juan y S de La
teoides. Esto sugiere que ambas especies podrían
tener un origen filogenético común, pero que ac-
tualmente estén aisladas geográficamente y posean
diferentes tipos de polinizador.
Material seleccionado. ARGENTINA, La Rioja: San
Martin, Sierra de U — falda E, frente a Ulapes. Hun-
ziker. A. y Caro 13562 (CORD): San Martín. Sierra de
арав, Cerro de Mona. Stuckert 17193 (CORD). San
jc Caucete, cerros al E de Marayes, Ruta 20, Múlgura
675 (SD: Valle Fértil, Quebrada de Mesada. Chucuma.
ы 738 (SI): Valle Fértil, 3 km al W de Maraves en
dec ción a San Agustín del Valle Fértil, Fortunato 5880
. Flor con bráctea. bractéolas, cáliz. corola. estambres у
(S1); Valle Fértil: de Marayes a Las Tumanas, Kiesling
3013 (S1). San е Ве Igrano, Sierra del Gigante, cerca
үтел 1936 (CORD): Бета,
Sierra del Sie en las i
Hunziker. A. у
— jones de La Calera.
Maldonado 16245 (CORD).
18. Justicia jujuyensis C. Ezcurra, Bol. Soc. Ar-
vent. Bot. 25(3—4): 350. 1988. Justicia leonar-
dii De Marco & Ruiz, Publ. Especial Inst. Li-
llo: 50. 1976. 1973. TIPO:
\rgentina. camino al Bana-
non Wasshausen
Jujuy: Ledesma.
nal. antes de llegar a Arroyo Quemado, en-
trando a la izquierda 3 km, leg. Legname y
Cuezzo 5970C (holotipo, LIL!
—
alt.. tallos
subcilindricos, cuadrisulcados, pubescentes a casi
Hierba sufruticosa de 0.5-1 n
=
con
elabros, con líneas longitudinales pilosas en su ju-
ventud. Hojas con pecíolo de 0.5-1.5 cm long. у
lámina ovada de 10-16 em long. por 1—7 cm lat.,
256
Annals of the
Missouri Botanical Garden
discolor, aguda y acuminada en el ápice, obtusa a
redondeada en la base, decurrente sobre el pecíolo,
con cistolitos poco notorios en su superficie, ge-
neralmente glabra en el envés. Flores sésiles dis-
puestas en espigas cortamente pedunculadas en las
axilas de las hojas superiores. Espigas más cortas
que las hojas, densas; brácteas florales lanceoladas,
de 8-10 mm long. por ca. 1 mm lat.
segmentos triangulares, de 3—4 mm long., agudos
glabros, ciliados en el margen.
to, patente.
long. y anteras con las tecas a distinta altura, |
superior de 2 mm long.,
mayor y con un apéndice glandular en la base. Es-
tilo glabro o pubérulo en la base; estigma bilobado;
ovario glabro y disco sinuado. Cápsula de ca. 1 em
long., con la mitad inferior sólida y lateralmente
comprimida y la superior subesférica y 4-seminada,
de 4 mm diám., glabra o pubérula. Semillas apla-
nadas, de contorno orbicular, de 3 mm diám., con
un reborde en su cara interna; retináculos obtusos,
de cerca de 2 mm long.
Ilustraciones. Lillo, 1937, lam. 2 (sub Justicia
nemoralis). De Marco y Ruiz, 1976: 51, 58 y 59
(sub J. leonardii). Ezcurra, 1993a: 357.
Distribución, hábitat y fenología. Sur de Boliv-
ia y norte de la Argentina, en las provincias de
Jujuy, Salta, Tucumán y Santiago del Estero. Habita
en lugares hámedos y sombreados de bosques de
transición en la zona de contacto entre las provin-
cias fitogeográficas Chaquefia y de las Yungas. Flo-
rece en verano, de enero a marzo.
Nombre vulgar. “Ichiyuyo” (Ezcurra, 1993a).
Justicia jujuyensis se asemeja mucho a J. corum-
bensis, pero habita una región más occidental y se
diferencia por las inflorescencias cortamente pe-
dunculadas más cortas que las hojas. Los nombres
Justicia nemoralis Lillo y Justicia nemoralis Lillo
var. tomentosa Lillo (Lillo, 1937) son nomina nuda
que fueron utilizados ampliamente en esta pn
cie en etiquetas de herbario. El nombre Justicic
nemoralis S. Moore (J. Bot. 47: 296. 1909) utilizado
para una especie de África es un homónimo ante-
rior que no permite su utilización para la especie
de América del Sur. Por esta razón se creó el nom-
bre nuevo Justicia jujuyensis para esta especie (Ez-
curra, 1988)
, foliáceas,
pubérulas; bractéolas lineal-lanceoladas, de 5-8
mm long. Cáliz 5-partido, de 4-5 mm long., con los
Corola blanca de
1.5 cm long., con tubo de 7 mm long. ensanchado
en una garganta acampanada de 5 mm diám., y
labios de 7-8 mm long., el posterior brevemente
bidentado, el anterior ampliamente trilobado erec-
Estambres con filamentos de 6 mm
p
la inferior oblicua, algo
Material seleccionado. ARGENTINA. Jujuy: Santa
Bárbara, Palma Sola a El Fuerte, ca. 5 km, Kiesling 5562
(SI); Ledesma, Calilegua, toma del Arroyo del Medio, Ca-
brera 31312 (SI); Valle G rande, Ledesma a Valle Grande,
Meyer 16416 (LIL). Salta: Orán, — io, camino a la
Cruz, Abbiatti y Claps 221 (LIL); Metán, Metán frente al
pueblo, 2 km W de la ruta 9/34, Novara 5826 (M, SI);
Rosario А la Frontera, El Naranjo, O'Donnell 5368 (LIL);
Rosario de — — — Venturi 8121 (K, SI).
— del Est Pellegrini, C° del eins, Ven-
turi 585: “Tu — teas . Oda. de la Hui-
guera, 55 la Sota 96 (LIL); Famaillá, Oda. de FA Río
Lules, Herrera 232 (LIL); Tafí del Valle, Yerba ICON
Venturi 88 (SI); Burruyacu, El Timbó, Venniri 2835 (LH
19. Justicia kuntzei Lindau, Bull. Herb. Boiss.
3: 483. 1895. TIPO: Bolivia. Santa Cruz: 1000
m, V 1892, Kuntze s.n. (B destr.; isotipo, NY!).
Lophothecium boliviense Bremek., Verh. Kon. Ned. Akad.
tensch. Afd. Naturk., Tweede sect. C, 72(4): 426.
1969. TIPO: Bolivia. — on kio dedi nis oil re-
finement and commencement of the pipeline, 2500
ft., 4 Sep. 1949, Brooke 5586 (holotipo, BM: isotipo,
NY!).
Sufrátice o arbusto de hasta 1 m alt. con ramas
cilíndricas, lisas, pubérulas en su juventud, gla-
brescentes y engrosadas por encima de los nudos
a la madurez. Hojas con pecíolo de hasta 1.5 em
long. y lámina elíptica de 6-12 em long. por 2.5—
5 cm lat. (la de las hojas apicales menor), aguda o
acuminada en el ápice, cuneada y decurrente en la
ase, esparcidamente pubérula en el haz y sobre la
nervadura en el envés. Flores sésiles, con 2 brac-
téolas en la base, en las axilas de brácteas lineales
y pequefias, agrupadas en espigas laterales reuni-
das en racimos simples o compuestos, laxos, pa-
niculiformes, generalmente axilares y más largos
que las hojas. Brácteas de las flores de 1.5-2.5 mm
long.; bractéolas menores. Cáliz de 3—4 mm long.
con los 5 segmentos lineales, agudos, glabros. Co-
rola lilacina manchada de blanco en la fauce, de
1.5-2 em long., con tubo basal de 3—4 mm long.,
garganta obcónica e inflada de 5 mm long. por 4
mm diám., labio posterior entero, de 10-12 mm
long. por 6-8 mm lat., el anterior trilobado, de 8—
10 mm long. por 5-6 mm lat., con lóbulos de cerca
de 1 mm long. por 2 mm lat. Filamentos de 10 mm
long.; anteras con las tecas superpuestas, separadas
por el conectivo de 1 mm long., la teca inferior
basalmente apendiculada, con un apéndice glan-
dular claviforme de 0.5 mm long. Estilo glabro; es-
tigma levemente engrosado; disco sinuado. Cápsu-
las de 1.5 em long. por 3—4 mm diám. con la mitad
inferior sólida у
lateralmente comprimida y la su-
perior engrosada, de contorno oblongo, estrechada
en su parte media. Semillas 4 o menos por aborto,
Volume 89, Number 2
2002
Ezcurra 257
Justicia en Sudamérica Austral
orbiculares, oscuras, de 1.5-2 mm diám., tuber-
culadas.
Ilustraciones. Ezcurra, 1993a: 355.
Distribución, hábitat y fenología. | Bolivia y N
a Argentina, donde habita en el sotobosque de
de
la selva en serranías de 600 a 1500 m de elevación.
Florece principalmente en primavera, de agosto a
noviembre.
Justicia kuntzei se caracteriza por las hojas gran-
des y las inflorescencias multifloras de flores me-
dianas con corola lilacina manchada de blanco en
la fauce
ARGENTINA. Jujuy: Lede "s-
o bosque alto, Venturi 5217 (SI, К);
40 (SI); Ledesma, 10 km NE
7. — Salta: Santa
2007 (SI); Anta,
Material seleccionado.
ma, Sa. Calilegua, baj
Capital, Zapla, Huan 306
of Ledesma, Eyerdam y Beetle 22275 (SI,
Victoria, PN Baritú, Arroyo Sidras, Brown 2
PN El Rey, camino a Pozo Verde, Novara y Charpin 10344
(LIL); Santa Victoria, camino de Baritú a — Mar-
mol 8734 (LIL); Orán, Tartagal, Schreiter 11370 (SI); San
Martín, Dique Itiyuro, Stange s.n. (LIL)
20. Justicia laevilinguis (Nees) Lindau, Engler
Bot. Jahrb. 19, Beibl. 48: 20. 1894. Rhyti-
glossa laevilinguis Nees, in Martius, Fl. Bras.
9: 120. 1847. Dianthera laevilinguis (Nees)
Durand & Jackson, Ind. Kew. Suppl. 1: 132.
1902. TIPO: Brasil. Rio Grande do Sul: Porto
Alegre, Sellow s.n. (sintipo, B destr.); San Ga-
briel, Estancia dos Fideles, Sellow s.n. (sinti-
po. В destr.); sin loc., Sellow s.n. (probables
isosintipos, K!, E!): Isla Santa Catarina, Twee-
die s.n. (sintipo, K no visto); Argentina. Cór-
doba: Río Segundo, Tweedie s.n. (sintipo, K!).
кее laevilinguis E s var. longifolia Nees, in Mar-
Fl. Bras. 9: 120. 1847. TIPO: Argentina. Bue
nos еи Rosle s.n. — G herb. DC no visto,
microficha!; probable isotipo. K!).
quon anagallis in Martius, Fl. Bras. 9: 119.
847.
Nees, 1
Justicia anagallis (Nees) — in Engler &
ae Nat. Pflanzenfam. 4(3B): 1895. TIPO:
Brasil. Bahia: Caitaité et Caxoeira, Mart ius s.n. (sin-
tipo, M!); entre Porto Alegre y Montevideo, L Ballons
s.n. (sintipo, В destr., fot. F 88081): — San
os llow s.n. (sintipo, B destr
Rhytiglossa repens Nees, in Martius, Fl. Bras. 9: 119.
1847. Ганна UND p or in Engler &
Р E Nat. Pflanzenfam. 4, Abt. 351. 1895.
Sellow
La
TIPO: Brasil. (бойро, " фан, fot. К
m T
Rhytiglossa obtusifolia Nees, in Martius, Fl. Bras. 9: 120.
isticia — (Nees) Lindau, Bull. Herb.
Boiss. 2, 1903. TIPO: Brasil. Rio
— do Sul: dia 2s gre; Sellow s.n. (sintipo, B
destr. Arg жашай Buenos Aires, herb. Arnott s.n.
(sint
"uius оннара Nees var. hirsuticaulis Nees,
artius, Fl. Bras. 9: 120. TIPO: Brasil. Río Negro,
Riedel s.n. (holotipo, LE no visto).
ше — S. Moore, Trans. Linn. Soc. 2, Bot.
132. )5. Justicia m (S. Moore) V. A. E
uin Kew Bull. 43(4): 600. 1988. TIPO: Brasil.
Moor re — eee бт Isotipos, МҮ! B destr..
fot. )!).
MEM a gaming Rusby, Mem. New York Bot. Gard.
. 19 TIPO: Bolivia. Rusby 1421 (holotipo,
Y!
Justicia ЕРИ —— Not. Mus. La Plata 13, Bot.
12: 92. 1948. TIPO: Argentina. Buenos Aires: Tigre,
D d = aM LP no visto).
Hierba palustre, con rizomas horizontales grue-
sos y tallos ascendentes, simples o ramosos, radi-
cantes en los nudos inferiores, subtetrágonos, cua-
drisulcados, de hasta 80 cm alt. y 5 mm diám.,
glabros, a veces algo pubescentes. Hojas superiores
sésiles. las inferiores con pecíolo de hasta 3 mm
long.. oblongas a lanceoladas, a veces lineales, de
3=12 (0.2—)0.5-1.5(-2.5) em lat.,
tusas en el ápice, redondeadas en la base, enteras
em long. y ob-
o levemente crenadas, generalmente glabras. Flores
en espigas unilaterales terminales o axilares de 5—
12 em long., con pedúnculos de hasta 8 cm long.
y entrenudos de ca. 1 cm long.: brácteas triangu-
lares de 1.5-3 mm long. por ca. 1 mm lat., y brac-
téolas triangular-lanceoladas de 1 mm long., ambas
tlabras. Cáliz 5-
ceolados de hasta 1 mm long. y 0.25 mm lat., agu-
1.5-2 em long. con tubo
partido con segmentos lineal-lan-
dos. Corola lilacina, de
basal de 6—10 mm long. por 3 mm diám., más an-
cha en la garganta: labio posterior ovado, de ca. !
cm long. y ca. 4 mm lat., obtuso a bidentado, v
labio inferior dá 13 mm long.,
lóbulos de 6 mm long. y ca. 5
algo más ancho y con dibujo pectinado-rugoso y
trilobado. con los
mm lat., el del medio
con manchas en la fauce. Estambres exertos, inclu-
sos debajo del labio posterior, con las anteras su-
perpuestas de menos de 2 mm long.. separadas por
el conectivo. Cápsula de ca. 2 cm long. y 6 mm
lat., estipitada, muy aplanada en la porción supe-
rior, apiculada y glabra. Retináculos de ca. З mm
Semilla suborbicular,
ong., bidentados. de ca. 5
mm diám., muy aplanada, glabra, con el margen
delgado y lacerado.
Ilustraciones. Dawson, 1979: 575.
Distribución, hábitat y fenología. Colombia,
Venezuela, Perú, Bolivia, sur de Brasil, Uruguay,
noreste de Argentina y Paraguay. Habita en am-
bientes palustres de regiones bajas. En el sur de
su distribución florece en primavera y verano, de
octubre a abril.
Justicia laevilinguis es una especie higrófila muy
variable en su morfología, con una distribución
muy amplia en América del Sur. Presenta especies
estrechamente relacionadas en América del Norte
258
Annals of the
Missouri Botanical Garden
como Justicia ovata (Walter) Lindau, que se ase-
meja en la morfología de tallo, inflorescencia y flor,
y cuyas relaciones deberían estudiarse.
ARGENTINA. Buenos Aires:
Material seleccionado.
Ribera — 1 del Río de la Plata, Hicken 12911 (SD;
La Plata, 7 — s.n. (К, SI). Corrientes: Saladas, Arbo
y Tur 2 22 214 (LIL); San Luis del Palmar, Ayo. Riachuelito,
25 Km E de * Luis del Palmar, Quarín 367 s AL); Con-
cepción, Ybarrola 324 (LIL); Gral. Paz, Gral. Paz, Ybarro-
la 3504 (LIL); а, RN 12, 20 km de н һаста
el E, Zuloaga 3230 . Chaco: Las Palmas, —
2335 (LIL); Resistencia, Margarita Belén, Meyer 379
(а T Puente de la Plaga, PN Chaco, Pajonal, Milgura
9 (SI). Entre Ríos: Uruguay, Concepción del Uru-
guay, Arroyo La C hina; campos vec inos, Bacigalupo 1376
SI); ¢ à i, Meyer 10281 (LIL).
Formosa: Pilcomayo, Estanci: ia Bouvier, riacho Araguay,
Guaglianone 454 (SI); Pirané, Pirané, Morel 161 (LIL).
isiones: Posadas, riberas del río, — ise 1168
(LIL); Coe. Pindapoy, Bertoni 243. s Santa
Fé: Gral. Obligado, Villa Ana, — vg IL); San
a Pto. Cabot. Morello
SI); Las Colonias, Esperanza, Ruiz Huidobro —
(LIL); Garay, Santa Rosa, Ruiz —* 3348 (LIL). I
AGUAY. Alto Paraguay: Alto „ Chaco, PURO
1269 (К, М); Руй Casado (С hac o) Кој 2199 (SI).
Amamb jay: in m in regione c oris f
Apa, Hassler 816€ ) (K, W), 8160a (BM. G,
regione cursus — шн — — 81604
(BM, G, NY). Ca zü: cerca y a ` de Yhú, junto al
arroyo Yhú, Fernández Casas et al. Р (С. МО, NY)
Canindeyú: in regione vicine — in regione Yerba-
lium de Marac :ayú, Hassler 4889 (BM, G, NY, P, W). Cen-
tral: L Assumption, sur le bord des marads, Balansa 2449
P); Man 5 del río Salado, camino de Limpio a Em-
Arbo et al. 1657 (CTES, ; Patino, near Asun-
м: Teague 572 (BM). Н on: Pr. Concepción,
ену» r 7624 (BM, G, NY), Hassler 7560 (BM, G, К, МО,
ЫР) Zw río Apa und río Aquidabán Centurión, Fiebrig
PA 30 (BM, E, G, K). Cordillera: Cordillera de Altos, Fie-
brig 356 (B ‚ К, M, US, W); Cordillera de Altos, in
campis, Hassler 3508 (BM, ( К, NY, Р, W); Cordillere
de Peribebuy, dans les ао s ——— Balansa 2451а
(G, Р); San Bernardino, costa del Lago Ipacaray, —
et al. 1510 (CTES, US); Tobaty, Arroyo Tobaty,
Rojas 4847 (US). Guairá: “ meadows, Villaric:
"d — (K, MO, NY,
K. P). N cembucú: ра Tene a la boca del río Ber-
mejo, Meyer 16102 (LIL). I ari: Loma Acahay, s.f.,
Chodat s.n. (G); Caballero, nius 435 (NY); Cerro Mba-
tobf, forest along N slope, in rege neration, Zardini y Ve-
lásquez 9935 ( 3M). Presidente Ha
164 de la ruta Trans e Schin, n
а 1031
‚ Mereles pris (C TE S. С).
San Pedro: in inen ida vicine San Estanislao, Hassler
1 { San Es stanislao, Rio Tapira-
cuay, Krapovickas et al. 13896 (CTES, US). S. «
ATA occidentalem flum. Paraguay, Gran Chaco, Rojas
930 (С).
Jerónimo, Riacho Coronda, frente
O36
In minis
=
=
‚ке
21. жн lilloana Ariza Espinar, Kurtziana 6:
1971. TIPO:
Fértil, campo de Ischialasta, 17
Kurtz 14223 (holotipo, CORD!).
Valle
1907.
San Juan:
Ene.
Argentina. $
Arbusto bajo, de hasta 0.50 m alt., con ramas
erectas, cilíndricas, las jóvenes verdosas y tetrá-
gonas y cuadrisulcadas, con estrías marcadas entre
las costillas, pubescentes en los nudos, de corteza
cenicienta a la madurez. Hojas cortamente pecio-
ladas, angostamente ovadas a elípticas, de 0.5-1.5
cm long. por 0.3-0.5 cm lat., obtusas, cuneadas en
a base, generalmente pubescentes. Flores sésiles
o cortamente pediceladas, con dos bractéolas en la
base, solitarias en las axilas de brácteas, formando
en conjunto espigas laxas; bractéolas lanceoladas
de cerca de 1 cm long. por 2 mm lat.; brácteas
obovadas de ca. 9 mm long. por 4 mm lat., ambas
glandulosas y ciliadas. Cáliz profundamente 5-par-
tido, con los lóbulos angostamente ovados a lan-
ceolados, subiguales, de 4 mm long., foliáceos, ge-
neralmente glabros. Corola azul violáceo, lilacina a
blanca de 2 cm long., pubérula, con tubo corto de
cerca de І em long., ensanchado en una garganta
de aproximadamente 0.8 em diám., los labios de 1—
1.3 em long., el anterior trilobado hasta la mitad,
con la fauce blanca, convexa y transversalmente
rugoso-venosa. Estambres insertos en la base de la
garganta, las anteras con las tecas subparalelas,
una por encima de la otra, la inferior generalmente
apendiculada en la base. Cápsulas claviformes,
muy fuertemente comprimidas, de 1.1—1.5 ст long.
por 5—7 mm ancho, glabras y de brillo nacarado en
su superficie, con el cuarto inferior sólido y late-
ralmente comprimido. Semillas generalmente 4, su-
borbiculares, aplanadas, de 6—7 mm diám., lisas;
retináculos obtusos.
Ariza-Espinar, 1971: 85.
Ilustraciones.
Distribución, hábitat y fenología. | Noroeste de
Tucumán y Catamarca
la Argentina, desde Salta,
hasta La Rioja y San Juan. Se la encuentra en re-
giones áridas entre 1300 y 2600 m s.m., en montes
xerófilos bajos, sobre suelo rocoso. Florece en ve-
rano, de diciembre a abril.
Nombre vulgar. “Alfalfillo” (Ariza-Espinar,
Justicia lilloana se caracteriza por sus hojas pe-
queñas y sus inflorescencias glandulosas, sus cáp-
y muy fuertemente comprimidas de
sulas grandes
y brillo nacarado, y sus semillas
superficie glabra
' lisas. El nombre Justicia
anchas, muy aplanadas y
platicarpa Lillo es un nomen nudum que se utilizó
para esta especie (Lillo, 1937).
Material seleccionado. ARGENTINA. Catamarca:
Belén, Los Nacimientos, Cabrera 16761 (CORD). La ۰
ja: Gral. Lavalle, Parque de Talampaya, Hunziker, J.
Gamerro 11669 (51); Independencia, e — Pagancillo y El
Balde (ruta Villa Unión a Patquía, km 96/97), Hunziker,
22441 (SI); Independencia, entre Puerta de Talampaya
Volume 89, Number 2
2002
Ezcurra 259
Justicia en Sudamérica Austral
Gualo, Kurtz 13260 (CORD). San Juan: Ischigualasto,
Roig 8053 (CORD): Iglesia. Cuesta del C^ Negro. entre
Palasto y Las Caldas Ruiz a Man (LIL). Tucu-
s dg Tafí. Amaicha, Diers 225 (SI); Trancas. Lomas de
Arcas. Schreiter 1105 (CORD. IL)
22. Justicia lilloi (J. L. Lotti) C. Ezcurra. in Ca-
brera. Fl. Prov. Jujuy (República Argentina) 9:
349. 1993. Chaetochlamys lilloi J. L. Lotti.
Publ. Especial Inst. Lillo: 63, fig. 1 y lám. |.
1976. TIPO: Argentina. Salta: Capital. 600 m
SMi;
Lillo 8087 (holotipo, LIL).
Hierba sufruticosa, de 0.50-1.50 m alt.. con ta-
llos subcilíndricos. cuadrisulcados, erectos. pubé-
rulos. Hojas con pecíolo de 0.5-2 cm long. y lám-
ina ovada o elíptica de 5-9 cm long. por 2-1 cm
lat., aguda en el ápice y cuneada o redondeada en
la base, verde clara, pubérula, con cistolitos den-
samente esparcidos en su superficie. las venas
principales prominentes en el envés. Flores sésiles
en dicasios condensados y generalmente reducidos
a una flor con dos bractéolas, agrupados en cortas
espigas con brácteas foliosas. Espigas axilares у
terminales, con el pedúnculo corto y el raquis re-
ducido a uno o dos entrenudos: brácteas muy si-
milares a las hojas, pero sésiles у algo menores. у
truncadas o cordadas en la base: bractéolas linea-
les. de hasta 15 mm long., pubescentes у corta-
mente ciliadas. Cáliz profundamente 5-partido. con
los segmentos ovado-lanceolados. de 10-
12 mm long. por 2-3 mm lat.. acuminados y densa
rígidos.
y cortamente ciliados, con el nervio medio promi-
nente en el envés. Corola lilacina de 34 em long..
con el tubo de 2 em long. por 3—4 mm diám.. poco
ensanchado en la garganta. los labios de cerca de
1.5 em long., el posterior de 7 mm lat.. bidentado.
el anterior profundamente trilobado. con los lóbulos
de 1.2 em long. por 6-8 mm lat. Estambres con
filamentos de 12 mm long. y anteras con las tecas
superpuestas y algo separadas por el conectivo. la
superior de 2.5 mm long.. la inferior de 3 mm long.
y apendiculada en la base. Estilo glabro con estig-
ma levemente engrosado y bilobado: ovario glabro
Cápsulas de
v disco en forma de copa. ondulado.
cerca de 1.5 em long. v 6 mm diam. con la mitad
inferior sólida y lateralmente comprimida y la su-
perior engrosada, ovoide y I-seminada. Semillas or-
lisas.
biculares subesféricas. de 2-3 mm diám.,
pardo-oscuras: retináculos de 3 mm long.
Ilustraciones. Lillo. 1937: lám. 7 y 8 (2) (sub
Chaetochlamys tucumanensis). Ezcurra. 1993a:
350.
Distribución. hábitat y fenología. Sudeste de
Bolivia. oeste de Paraguay y norte de la Argentina.
Habita principalmente en facies húmedas y bajas
del Chaco Serrano septentrional y en bosques de
transición a las selvas montanas de las Yungas en-
tre 500 y
cuentra en laderas de cerros del Chaco Paraguayo.
1200 m de elevación. También se en-
Florece de fines de primavera a otoño, de diciembre
a mayo
Chaetochlamys tucumanensis Lillo es un nomen
nudum con el que originalmente se denominó i
sla especie (Lillo, 1937: Lotti, 1976). Justicia lilloi
es una especie muy afin a Justicia thunbergioides
(Lindau) Leonard, descripta para Mato Grosso do
Sul. Brasil (Brasil. Mato Grosso pr. Corumba. Mal-
me 3026. sintipo В destr.. fot. F 89191).
mites entre ambas no son claros (Ezcurra, "1993b).
v los lí-
Junto con Justicia allocota Leonard (Brasil. Goiás:
Yale Dawson 15023, holotipo, US!) y Justicia ve-
nusta (Rizzini) V. A. W. Graham (Brasil. Minas Ge-
Brade 17563, isotipo. US!) forman un grupo
de diferenciación compleja que parecen todas de-
rais:
rivadas de un antecesor común y adaptadas a di-
ferentes regiones con distintas condiciones de hu-
medad y precipitación. Justicia lilloi es la que
presenta generalmente hojas más chicas y se en-
cuentra principalmente en las selvas pedemontanas
de las Yungas de Bolivia y Argentina y en serranías
del Chaco Paraguayo, mientras que Justicia thun-
bergioides y Justicia allocota se encuentran en re-
giones más húmedas del sudoeste y centro de Bra-
sil respectivamente.
Material seleccionado. ARGENTINA. Catamarca: El
Alto. entre Alijilán y El Alto, Cristóbal 340 (LIL) нута
Santa — Abra de los Morteros. Cabrera 21719 (S
San Pedro, San Pedro a El Milagro entrando hacia 2
M лени 5372 (LIL). Salta: Rosario de la Frontera,
halle 4417 (LIL): Ор he Vie 03.
Lomas al N
— $532 (SI): Orán, Cerro Tartagal. /
: Metán. Metán. Roc ha 108 (LIL J
[iners Toledo 21
200 m antes de Vipos, | Vereoorst 1221 (SI): (
Barranca Colorada. Venturi 1150
de la finca Guillermina,
camino a — Medina, Villa Carenzo 1750 (V.
AGL "araguay: Cerro Pedrera. Rio Navileque.
Fuerte Olimpo, — 20425 (С); Fuerte Olimpo, Cha-
s 13842 (LIL): Fuerte Olimpo. in co-
Е — Bernardi 20335 (©.
desde lomada al S (cam-
co Paraguayo.
linis Tres Marías.
MO). Chaco: Chaco. Cerro León,
ы е. hasta mesela J entral, Charpin y Ramella 2173
(P)
p-p- (G). S. dep.: Paraguay. "кады ALIZ
23. Justicia lythroides (Nees) V. A. W. Graham.
Kew Bull. 43: 603. 1988. Heinzelia lythroides
Nees. in Martius, Fl. Bras. 9: 154. 1847.
TIPO: Brasil. Rio de Janeiro: 7
rahy.” Pohl s.n. (holotipo. W!).
ad fluvium Pi-
Heinzelia ovalis Nees. in Martius. Fl. Bras. 9: 154. 18
260
Annals of the
Missouri Botanical Garden
non — ovalis Ridley, J. Fed. Malay States Mus.
10: 150. 1820. TIPO: Brasil. Sáo Paulo: Sáo Paulo.
Ypanema, in d Riedel 1983 (lectotipo, aquí de-
signado, LE, fotogr. US!).
Hierba perenne, rizomatosa en la base, con tallos
tenues, subcilfndricos, decumbentes y geniculados,
luego erectos, de hasta 40-70) em long., pubéru-
los. Hojas opuestas con pecíolo de 3-10 mm long..
y lámina elíptica a angostamente ovada de 3—8 cm
long. por 1.3-3 em lat., aguda o acuminada en el
ápice, cuneada y algo decurrente en la base, entera,
generalmente glabra, a veces pubérula sobre el ner-
vio medio en el haz, verde clara y pubescente sobre
la nervadura en el envés. Inflorescencias formadas
por flores en espigas unilaterales tenues, simples o
compuestas y ramificadas, de 4-8 cm long.. ter-
minales o en las axilas de las hojas superiores, cor-
tamente pedunculadas, con el eje arquado. Brác-
teas y bractéolas lineal-lanceoladas, rígidas. de 1—
3 mm long., pubérulas, con el nervio medio pro-
minente en el dorso. Cáliz de cerca de 6 mm long..
con los 4 segmentos de 5 mm long., rígidos, linea-
les, subulados, dorsalmente pubescentes sobre el
nervio medio. Corola blanca, pálido lilacina o vio-
lácea de ca. 1 em long., con tubo de 0.5-0.6 cm
long. por 1.5 mm diám., ampliado en la garganta,
el labio superior erecto, lanceolado, de 3 mm long.,
agudo, el inferior ancho, trilobado, de ca. 4 mm
long., con los lóbulos de ca. 2 mm long., redon-
deados. Estambres con filamentos de aprox. 4 mm
long.; teca apical perfecta, de 1 mm long., la in-
ferior estéril, reducida a un leve engrosamiento so-
bre el filamento. Cápsula de contorno angostamente
elíptico, de aprox. 5 mm long. por 1.5 mm lat.,
cortamente sólido-estipitada en la base, glabrescen-
te y de paredes finas. Semillas 4, algo aplanadas,
de 1 mm diám., pardo-oscuras, con pelos cortos y
gloquidiados; retináculos de | mm long., obtusos.
Ilustraciones. Nees, 18; (a:
Acosta-Castellanos, 1907:
tab. 97. Ezcurra y
Distribución, hábitat y fenología. Sur de Brasil.
noreste de Argentina y Paraguay oriental. Habita
en el sotobosque y en bordes de bosques htimedos,
frecuentemente en lugares alterados. Florece en
otofio e invierno, de abril a agosto.
Justicia lythroides se caractriza por sus flores en
espigas secundifloras tenues y delicadas y sus flo-
res pequefias. Es muy afín a la especie descripta
bajo el nombre de Chaetothylax hatsbachii Leonard
(Brasil. Paraná: Cerro Azul, Turvo, Rio Ribeira,
Hatschbach 5419, holotipo US!), que podría ser si-
nónimo de esta especie.
El nombre Heinzelia ovalis fue creado por Nees
sobre la base de tres ejemplares: dos de Brasil (Sâo
Paulo, Ipanema, Riedel 1983, y Minas Gerais, Bar-
bacena, Riedel 88, LE) y uno de Argentina, Tucu-
man (Tweedie s.n., K!). Estos sintipos comprenden
una mezcla de dos especies. El ejemplar coleccio-
nado en Argentina por Tweedie pertenece a Justicia
goudotii, mientras que el material de Brasil colec-
cionado por Riedel, del que he visto una —
y un probable isosintipo (Brasil. Sáo Paulo,
sylvula ad vias pr. Ypanema, 111-1834, Riedel s.n.,
NY!) pertenece a Justicia lythroides. Debido a que
a diagnosis de Heinzelia ovalis coincide más con
as características del material de Riedel que con
el de Tweedie en tamaño de hojas y flores, designo
como lectotipo de Heinzelia ovalis al ejemplar de
Justicia lythroides de Brasil, Ipanema, Riedel 1983
(LE). Esta designación hace que el nombre Hein-
zelia ovalis deba considerarse sinónimo de Justicia
lythroides.
Material seleccionado. ARGENTINA, Misiones: El-
dorado, Eldorado, km 28, Burkart 14603 (SI); Frontera,
reforestaci ‘ión Gral. М
dez 144 (CTES): Cainguás, Arroyo Caña-Pirú, 8 km de A
del Valle hacia Jardín de América, Hunziker, J. 10891
SI); San Ignacio, Ayo. Macaco, Schwarz 5687 (LIL); Mon-
; 3‹ :
a 2188 (SI)
Alo: Par araná Fer 6040 (G); in regione fluminis Alto
Parana, Fisbrig 6173 (G, SI); Río Itabó, sotobosque, ribera
n río, пай к 1001 (МО). Caazapa: Tavaf,
Castor Cue, S, 55°20'W, Degen 1611 (MO); Tavat,
Compañía ко. — 1522 (G, MO), Ortiz 1280 (G,
MO). Canindeyú: Mbaracayú, comunidad Mbyá, Basua-
lado 2546 (FCQ).
24. Justicia mandoni (Lindau) Wasshausen & C.
Ezcurra, Candollea 52: 175. 1997. Beloperone
mandoni Lindau. Bull. Herb. 5: 615.
1897. TIPO: Bolivia. La Paz: Larecaja, in vi-
cinity of Sorata, Cerro Iminapi, Mandon 297
(holotipo, B destr.; isotipos, BM! p.p., K! p.p.).
Boiss.
Justicia — De * 0 & Ruiz, Publ. Espec. Inst.
Lillo: 53, fig. 3 y láms. 5 y 6, 1976. TIPO: n
ш. Chicli e) Puesto Las Pavas, 1060
1., Meyer 15125 (holotipo, LIL!).
Sufrútice o arbusto apoyante de 1-2 m alt., con
ramas cilíndricas, lisas o muy finamente estriadas,
generalmente glabras. Hojas con pecíolo de 0.5-1
em long., pubérulo, y lámina angostamente ovada,
de 4-8 cm long. por 1.2-2.5 em lat., redondeada y
levemente decurrente en la base, acuminada en el
ápice, levemente pubescente en el envés, pubérula
en el haz, con nervaduras marcadas. Flores sésiles
con dos bractéolas basales en las axilas de brácteas
dispuestas en espigas generalmente terminales, las
brácteas opuestas pero alternadamente anisofloras,
dándole un aspecto unilateral a la espiga. Brácteas
Volume 89, Number 2
2002
Ezcurra 261
Justicia en Sudamérica Austral
2.0
foliosas, pubescentes: bractéolas lineal-
elíptico-lanceoladas, de 1-1.5 em long. por 1.5—
mm lat.,
lanceoladas, de 1-2 mm lat., algo más cortas que
el cáliz. Cáliz profundamente 5-partido, los seg-
mentos — ovado-lanceolados, de 1.5-2 ст
long. por 2-3 mm lat., generalmente glabros a pu-
bérulos. — rojo-morada de 3-3.5 em long.. el
tubo de aprox. la misma longitud que los labios v
de 5 mm diám.. interiormente surcado en la parte
posterior por dos pliegues longitudinales que alojan
el estilo, el labio posterior bidentado, de 1.5-2 cm
long. por aprox. 8 mm lat., el anterior trilobado, de
7 mm lat., reticulado-venoso y convexo en la base.
con los lóbulos de aprox. | em long. por 3—4 mm
lat. Estambres con filamentos de 2 em long. v tecas
situadas a distinta altura, la superior de 2 mm
long., mutica. la inferior separada del conectivo.
algo más grande y apendiculada en la base. Cáp-
sula de contorno oblongo, de 1.5-2 cm long.
0.5
aplanadas de 2 mm diám., verrugosas; retináculos
por
em didm., pubescente. estipitada. Semillas
de 2 mm long.
Ilustraciones. Lillo. 1937: lam. 6 (sub Justicia
tucumanensis). De Marco y Ruiz, 1976: 54. 60 y
Ol (sub Justicia odonellit). Ezcurra, 1993a: 337
(sub Justicia odonellii).
Distribución, hábitat y fenología. Sudeste de
Bolivia y Norte de la Argentina, en las provincias
de Jujuy. Salta v Tucumán. Habita en las selvas de
la provincia fitogeográfica de las Yungas en serra-
nfas entre 1000 y 1700 m, generalmente apovada
sobre otros arbustos o árboles. Florece en otoño.
invierno y primavera, de febrero a septiembre.
Justicia mandoni es un arbusto apoyante de sel-
va que se caracteriza por las grandes flores rojo-
moradas en espigas de aspecto unilateral. Se ase-
meja a J. dumetorum, pero esta última es de porte
menos elevado, sus inflorescencias no son unilate-
rales, y sus flores tienen labios más cortos en re-
lación a la longitud del tubo basal. Por un error
involuntario, recientemente se trató a J. mandoni
como especie endémica de Argentina (Ezcurra.
1999), a pesar de que su distribución abarca el sur
de Bolivia. Justicia tucumanensis es un nomen nu-
dum con el que originalmente se denominó a esta
especie en Argentina (Lillo, 1937).
ARGENTINA, Jujuy: Capital,
Material selece we
le antena, Cabrera
Хара. Mina 9 de Octubre,
5 Calde ra, ca. Abra de
Ruta 9. km 1
> Abra de Santa Laura, Novara
8055 (M. 51); Santa p Чопа, Camino de Toldos al Lipeo,
a. 20 km de Toldo ате y Te 8163 (LIL): Anta,
campo Las Heras, — 966 (LIL); Rosario de la Fron-
Venturi 9338 (К). а Tafí, quebrada de los
subida a la
brera 3413. Caldera,
N de Ojo de с 5 km <
^
lera,
Arroyo Azucena, 10 1079 (St): Chicligasta.
Piscicultura, Meyer 13853 (LIL): Barravacu, Barravacu,
Türpe 40 (LIL); le y Villa Nogués, Venturi 1799 (SI).
Sosa,
25. Justicia oblonga (Nees) Lindau. in Engler &
Prantl, Nat. (3b): 350. 1895.
atada oblonga Nees, in Martius, Fl.
Bras. 9: 124. 1847. TIPO: Brasil. Rio Grande
do Sul: Porto Alegre, Sellow s.n
destr.. fot. F 8847):
San Lorenzo, Sellow s.n. (sintipo, B destr.): sin
Pflanzenfam.
(sintipo. B
Uruguay. Santa Lucía у
loc., Sellow s.n. (probable isosintipo, K!).
Hierba perenne, rizomatosa en la base, con tallos
erectos, subcilindricos, de hasta 60 em long..
bérulos. Hojas opuestas con pecíolo de 0.2
long. y lámina oblongo-elíptica de 2-5 em lace por
].2—9
a
cm lat., acuminada en el ápice, cuneada
algo decurrente en la base, entera, coriácea, pu-
bérula en el haz, pubescente sobre la nervadura en
el envés. Inflorescencias formadas por espigas den-
—
sas, sésiles, terminales o en las axilas de las hojas
superiores, a veces formando racimos de espigas en
la porción terminal de los tallos. con las hojas re-
ducidas. Brácteas lanceoladas. rígidas. de 5-7 mm
long.. pubérulas, ciliadas. Cáliz de cerca de 5 mm
long.. con los 4 segmentos rígidos. lineales. subu-
lados, dorsalmente pubérulos, ciliados. Corola pá-
lido lilacina a violácea de ca. | em long., con tubo
de 0.5-0.6 em long., el labio superior erecto, de
3.5 mm long. por | mm lat., el labio inferior ancho.
trilobado, de ca. 3 mm long., con estrías blancas
en la fauce. Estambres con filamentos de aprox. 2
mm long.: teca apical perfecta, de contorno ovado.
de ca. | mm long., la otra inserta más abajo en el
filamento, más pequeña, angosta y basalmente cau-
dada. Cápsula de contorno obovado, de aprox. 8
mm long. por З mm lat., cortamente sólido-estipi-
tada en la base, glabrescente y de paredes finas.
Semillas 4,
eloquidiados: retináculos de ca.
de ca. | mm diám., con pelos cortos у
| mm long., se
miobtusos.
Ilustraciones. Ezcurra y Acosta Castellanos.
1997: 109
Distribución, hábitat y fenología. Sur de Brasil.
Uruguay, noreste de Argentina y Paraguay oriental.
Habita en ambientes húmedos. Florece principal-
mente en verano y otoño, de noviembre a abril.
Justicia oblonga se caracteriza por sus flores de
aprox. | em long., en espigas densas con brácteas
pequeñas. Por las características de sus inflores-
cencias, anteras y sus cápsulas y semillas induda-
blemente pertenece a Justicia sect. Chaetothylax
Nees) V. A. W. Graham (Tabla
sus corolas poseen un tubo corto, lo que no es muy
1). a pesar de que
—
262
Annals of the
Missouri Botanical Garden
frecuente en esta sección (Ezcurra & Acosta-Cas-
tellanos, 1997).
Material se s ccionado. ARGENTINA. Misiones:
campos, Burkart 14443 (Sl); Can-
odriguez 288 (Sl); San Javier, San
Javier, Krapovickas 15161 (CTES, SI); Leandro
Paso Carreta, Krapovickas 15021 (CTES, SI); Candelaria,
Berón Astrada, Yabebyry, Montes 841 (SI); San Ignacio,
ico, Schwarz 5706 (LIL); Cainguás, ruta 1 ^i Km
952. : — 3065 (LIL). Corrientes: Ituzaingó, |
z de Ruta 12, camino a San Carlos,
CTES); Santo Tomé, Estancia Garruchos, potrero Puente,
Krapovickas 21611 (CTES): Santo Tomé, Ea. Timbó, Avo.
Ciriaco у Ruta 40, Schinini 23450 (CTES, LIL). PARA-
GUAY. Alto Parana: in — fluminis Alto Parana,
Fiebrig 5929 (BM, G, K, LIL, SI); Хас iei Montes 9806
; Nacunday, Mantes 1093 3 (L P). Guaira: от
Montes 16162 (LIL); Iti — Monies 15762 (LIL): Yegua
Porá, Montes 16383 (LIL
>?
E
2
a
—
>
=
S
©
ч.
=
=
=
a
>
co
m
—
—
26. Justicia oranensis De Marco & Ruiz, Publ.
Especial Inst. Lillo: 46, fig. 1 y láms. 1 v 2,
1976. TIPO: Argentina. $
la Finca de Jakulica, a 8 km del puente del
Salta: Oran, camino a
Rio Bermejo, 600 m s.n.m., Legname y Cuezzo
8564c (holotipo, LIL).
Arbusto ramoso de 1—3 m alt.. con ramas sub-
cilíndricas, las jóvenes cuadrisulcadas, con dos
bandas longitudinales pilosas a lo largo de los en-
trenudos. Hojas con pecíolo de 1—4 cm long. y lám-
ina ovada o elíptica de 5—12 cm long. por 3-8 em
lat., aguda y levemente acuminada en el ápice, cu-
neada en la base, parcialmente decurrente sobre el
pecíolo, con las venas principales marcadamente
prominentes y notorias en el envés, las secundarias
arqueadas y paralelas, la nervadura en general pu-
bescente. Flores sésiles dispuestas en espigas den-
sas y bracteadas que se agrupan en racimos o pa-
nojas laxos, pedunculados, más largos que las
hojas, en las axilas de las hojas superiores. Brác-
teas florales obovadas, de 5-8 mm long. por 3-5
mm lat., redondeadas, corta y densamente glandu-
loso-pubescentes; bractéolas oblongas, de aprox. 6
mm long. por 2 mm lat., obtusas, glandulosas. Cáliz
profundamente 5-partido, con los segmentos lan-
ceolados, de 6-7 mm long. por 1-2 mm lat.. agu-
dos, levemente glandulosos. Corola rojo-anaranja-
da, de cerca de 2.5 cm long., con tubo de 1.2 cm
long.. ensanchado а 3—4 mm diám. en la garganta,
y labios de 1 cm long., el posterior bidentado, el
anterior trilobado, con los lóbulos de 1 mm long.
Estambres con filamentos de 1.5 cm long., y anteras
con tecas a distinta altura: la superior separada por
el conectivo y oblicua, de 2 mm long., la inferior
algo más larga, apendiculada en la base. Cápsula
de 1.3-1.5 em long., estrechada y sólida en la mi-
tad inferior, la porción superior inflada y 4-semi-
nada, glanduloso-pubérula. Semillas suborbiculares
de 3 mm diám., tuberculadas. Jaculatores de 2 mm
long.
Ilustraciones. De Marco y Ruiz, 1976: 47, 56 y
57. Ezcurra, 1993a: 341.
Sur de Boliv-
Argentina, en las provincias de
Distribución, hábitat y fenología.
ia y norte de la
Jujuy, Salta y Tucumán. Habita en selvas bajas de
la provincia fitogeográfica de las Yungas entre 500
y 1600 m, y florece en invierno y primavera, de
julio a noviembre. Especie llamativa por sus pa-
nojas de flores rojas, según De Marco y Ruiz (1976)
es fácilmente cultivable y se puede propagar por
estacas.
Justicia oranensis es una especie llamativa por
sus panojas de espigas de flores rojo-anaranjadas,
y según De Marco y Ruiz (1976) es fácilmente cul-
tivable y se puede propagar por estacas. Justicia
schreiteri Lillo es un nomen nudum que se utilizó
para esta especie (Lillo, 1937).
ARGENTINA. Jujuy: Ledes-
— та 27954 (SI); Ledesma, PN Cali-
9 (SI). Salta: Santa Victoria, camino
de Los Toldos al Lipeo, Legname y Cuezzo 9707 (LIL);
Tabacal, Martinez Cet 3376 (LIL ); — José de San
Martín, Tartagal, 5 1, Cedral, Ro-
driguez 1045 (LiL): — Vado ное жене 7043 (
(1
Ма terial seleccionado.
ma, Caimancito,
legua, Cabrera 3214
AL). Tucumán: sin loc., Baer 80 (SI).
27. Justicia pectoralis Jacquin, Enum. Syst. Pl.
11. 1760. Dianthera pectoralis (Jacq.) Murr.,
Syst. Veg. Ed. 14, 64. 1784. Stethoma pecto-
ralis (Jacq.) Raf., Fl. Tellur. 4: 61. 1838
(1836). Rhytiglossa pectoralis (Jacq.) Nees, in
London J. Bot. 4: 637. 1845. Ecbo-
lium pectorale (Jacq.) Kuntze, Revis. Gen. РІ.
2: 487. 18
Bentham,
1 . Psacadocalymma pectorale
(Jacq.) Bremek., Verh. Kon. Ned. Akad. We-
Afd. Natuurk., 45: 55.
1948. TIPO: Ilustración en Jacquin, Selec.
Stirp. Hist.,
designado ante la ausencia de ejemplar tipo,
1995).
tensch., Tweede Sect.
Amer. t. 3. 1763 (lectotipo, aquí
ver Daniel,
Hierba pequeña, con tallos ascendentes, radi-
cantes en los nudos inferiores, de hasta 30 cm alt.
y | mm diám., pubescentes. Hojas con pecíolo de
3—4 mm long., lanceoladas a ovado-lanceoladas, de
4-6 cm long. y 1-2 em lat., agudas en el ápice,
redondeadas en la base, enteras o levemente cre-
nadas, levemente pubescentes sobre las venas. Flo-
res en espigas unilaterales terminales y axilares,
simples (en el material de Argentina y Paraguay) o
ramificadas, a veces agrupadas en inflorescencias
paniculiformes, de 4-10 cm long., con pedúnculos
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
de 1.5-2.5 em long. y entrenudos de menos de 1
¢
em long.; brácteas lanceoladas de 2.54 mm long.
por 0.5 mm lat., y bractéolas lanceoladas de 2 mm
Cáliz 4-partido con seg-
mentos de 3.5 mm long. y 0.25 mm lat.,
pubérulos. Corola blanca de 0.8-1.2 cm long. con
long.. ambas pubérulas. (
obtusos,
tubo basal de 5 mm long. por 2.5 mm diám. en la
fauce: labio posterior ovado, de 3 mm long. y ca. 2
mm lat., obtuso. y labio inferior de 5 mm long.,
trilobado, con los lóbulos de 3 mm long. y ca. 2
mm lat., el del medio algo más ancho y con man-
chas rosadas en la fauce. Estambres exertos, inclu-
sos debajo del labio posterior. Cápsula de ca. Т cm
long., superiormente inflada y 4-seminada. Semillas
orbicular-oblongas, aplanadas, de ca. 1 mm diám.,
pusticuladas, pardo-oscuras.
Durkee, 1986: 12.
Ilustraciones.
Distribución, hábitat y fenología. América trop-
ical hasta el Sur de Brasil, y Paraguay oriental y
regiones limítrofes de Argentina. En el sur de su
distribución habita en el sotobosque de bosques es-
tacionalmente secos, generalmente sobre suelo are-
noso. Florece en otoño y primavera, de septiembre
a abril.
El material de Justicia pectoralis de Paraguay y
Argentina se diferencia de la mayoría de los ejem-
plares de esta especie que habitan más al norte por
el porte bajo y las espigas terminales generalmente
simples y no agrupadas en inflorescencias panicu-
liformes. Esta forma empobrecida podría deberse al
efecto de una temporada favorable más corta por
efecto de su crecimiento a una mayor latitud geo-
eráfica, con veranos de menor duración. Esta mor-
fología se asemeja a la descripción de Justicia sar-
(Nees) Lindau basada en Rhytiglossa
Sellow 56, holoti-
mentosa
sarmentosa Nees (Brasil. S. loc.,
po, B destr.). La ausencia de material tipo asociado
a este epíteto hace dudosa su identificación.
ARGENTIN A. Formosa: Ma-
Material selece —
tacos, Ing. Juarez, chacra mataca, 3 km camino al R. Ber-
mejo, Maranta 295 | SI): Ing. Juarez Toldería Toba. 1 km
al N del pueblo, Arenas 2283 (CORD, SI). PARAGUAY.
Alto INE San Carlos del Alto Paraguay, Molas
). Boquerón: Chaco Paraguayo, ruta Trans Cha-
со. Schinini y Palacios 25788 (CTES). Chaco: Mayor Pe-
dro Lagerenza, selva ber del Río Timane, Schinini y
Bordas 14938 (CTES, С). ieva Asunción: Teniente
Primero A. Picco, Spic — et we 2298 (G).
28. Justicia phyllocalyx (Lindau) Wasshausen &
Poikt-
Lindau, Engl. Bot.
1989. TIPO: Brasil.
Cocal, Glaziou
Ezcurra, Candollea 52: 175. |
e
lacanthus phyllocalyx 1
Jahrb. 25. Beibl. 60: 48.
Goyaz: inter As Brancas et
21876 (holotipo. B destr..
K!).
fot. F 87851; isotipo.
— cien macedoana um TÉ Jard. Bot. Rio Ja-
iro 8: 357, t. 7 y 8, f. 3. . TIPO: Brasil. Minas
— Uberlandia, el (holotipo. RB fot.
US».
СЕК тас — Rizzini var. — — for-
ma redacta Rizzi Dusenia 3: 189. TIPO:
ras al. inas Cau Ituiutaba, — — (ho-
lotipo, RB: isotipo, US!).
Hierba erecta, de 20-70 em alt. con rizoma en
a base, y tallos poco ramificados, cilíndricos, gla-
bros a pubescentes o cortamente tomentosos. Hojas
cortamente pecioladas, anchamente elípticas, de 3—
6(-8) em long. y 1.5-3 cm lat., agudas, cuneadas.
glabras a densamente pubérulas o cortamente to-
mentosas. Inflorescencias en espigas terminales o
axilares de 3-8 cm long., sésiles a cortamente pe-
dunculadas, alternadamente anisofloras (una sola
de cada par de brácteas portando una flor con sus
dos bractéolas en la axila); brácteas lineal-lanceo-
ladas, de 10-15 mm, más o menos glanduloso-pu-
bérulas y seríceas; bractéolas lineales, de 8-10 mm
long., pubérulas o glabras. Cáliz profundamente 5-
partido, con los segmentos subiguales, elíptico-ova-
dos, de 1.2-1.5 em long. por 3—4 mm lat.. agudos
o acuminados, pubérulos o glabros, trinervados, con
el nervio medio marcado. Corola blanquecina, de
ca. 2 em long., con el tubo de 1.3 cm long., acam-
panado, hasta de 0.7 ст diám. en la fauce, el ló-
bulo posterior cóncavo, anchamente ovado, de ca.
у 0.7
trilobado, reticulado-venoso en la fauce, con los ló-
| em long. em lat., bidentado, el anterior
bulos anchamente oblongos, de ca. 1 em long. y 8
mm lat., redondeados. Estambres exertos por de-
bajo del labio posterior, las tecas marcadamente su-
perpuestas, separadas por un conectivo ancho. la
inferior apendiculada en la base. Cápsula desco-
nocida.
Ilustraciones. Figura 11.
Distribución, hábitat y fenología. Sudoeste de
Brasil, y noreste de Paraguav en la región del
Amambay. Habita en cerrados, en bordes de
bosque. Florece en verano y otoño, de enero a abril.
Justicia phyllocalyx se asemeja a otras especies
de Justicia sect. Simonisia como J. lilloi v J. rusbyi,
por lo que se la ubica en esta sección (Tabla 1). Al
igual que estas especies, posee polen biporado,
pero соп ínsulas sobre toda su superficie, no sola-
mente en hileras a los costados de los poros. En
esto se asemeja a especies de Poikilacanthus. El
polen 4- u 8-porado y la exina con ínsulas en toda
su superficie que le dan un aspecto facetado es
característico (Daniel. 1991.
1998), pero el polen de esta especie no coincide
de Poikilacanthus
totalmente con el de Potkilacanthus. A pesar de las
ínsulas en toda su superficie, debe ser tratado en
264
Annals of the
Missouri Botanical Garden
Figura 11.
Justicia por poseer 2 poros y por la disposición de
las ínsulas, y por eso fue transferido a este género
(Wasshausen & Ezcurra, 1997).
Material seleccionado. 'ARAGUAY. Amambay: Ar-
a (G); in re — cursus superioris fluminis
Apa, Hassler 8423 (BM, G, K, LIL, , W); Parque
Nacional Cerro Corá, open — NE of lie 'adquarters,
Solomon 7055 (MO, PY), 7056 (MO); Sierra de Amam-
bay, ad marginem silvarum pr. m lla, Rojas 10152 (G,
MO).
Justicia phyllocalyx. —A. Rama con flores.
—D. Corola abierta con androceo. —E. Cápsula abierta sin semillas. Hassler 8423 (BM, MO),
(51)
—C. Corola.
be urra 1831, 1840
—B. Nudo con bráctea, bractéolas у
—
29. э polygaloides (S. Moore) Lindau,
Jull. Herb. Boiss. Ser. 2, 3: 633. 1903. Dian-
кук, polygaloides S. Moore, Trans. Linn. Soc.
Ser. 2, 4: 433. 1895. TIPO: Brasil. Mato Gros-
so: Santa Cruz, Moore 667 (holotipo, K!; iso-
tipo, B destr., fot. F 8855!).
Hierba pequefia, con tallos ascendentes, radi-
cantes en los nudos inferiores, de hasta 30 cm alt.
E
y 1.5 mm diám., glabros o pubérulos. Hojas sub-
sésiles o cortamente pecioladas, lanceoladas o an-
Volume 89, Number 2
2002
gostamente oblongas, de 3-6 em long. y 0.4-1 cm
at., agudas en el ápice, cuneadas en la base, en-
teras, generalmente glabras. Flores en espigas ter-
minales simples, de 3-8 em long.. con pedúnculos
de 1.54 em long. y entrenudos de menos de | cm
long.: brácteas lanceoladas de 1.5—4 mm long. por
0.5 mm lat., y bractéolas triangulares de 2-3 mm
long., glabras o pubérulas. Cáliz 4-partido con seg-
mentos de 3-5 mm long. y 0.25 mm lat.. acrescen-
te. Corola lilacina de 1 em long. con tubo basal de
5 mm long. por 2.5 mm diám. en la fauce: labio
ca. 2 mm lat..
posterior ovado. de 3 mm long. v
obtuso, v labio inferior de 3 mm long.. trilobado.
con los lóbulos de 2 mm long. y ca. 2 mm lat., el
del medio algo más ancho y con dibujo pectinado-
rugoso, blanco. en la fauce. Estambres exertos. in-
clusos debajo del labio posterior, con las anteras
con las tecas superpuestas de ca. 1 mm long.. la
inferior oblicua. Cápsula de contorno elíptico. de
0.7 mm long. y 2 mm grosor, glabra. Semillas 4.
aplanadas, de ca. | mm diám., diminutamente ve-
rrucosas.
Figura 12.
Ilustraciones.
Distribución, hábitat y fenología. Sudoeste de
Brasil y Paraguay oriental. Habita frecuentemente
en prados ribereños. Florece en primavera y vera-
no, de septiembre a febrero.
Justicia polygaloides se caracteriza por las hojas
lanceoladas o angostamente oblongas y las flores
de aprox. | em long. en espigas terminales simples.
Es una especie relativamente rara, que ha sido co-
leccionada pocas veces, en general en prados ri-
bereños. El ejemplar de Argentina, Entre Ríos. pre-
senta una distribución anómala, pero se asemeja al
resto del material.
seleccionado. ARGENTINA. Ríos:
alneario e Tortuga Alegre, al N de Concor-
dia, li 997 AGUAY. Alto dorm
Puert to Casado, Rojas ne (SI). € oncepción: Pr. Con-
. Hassler 7653 (BM, G,
laterial Entre
=
>
=ч
“~~
layes: Gran Chaco, Loma Clavel. Rojas 2631 (BM. G
К. NY): in regione cursus inferioris fluminis Pilcomayo,
T 669 (G). San Pedro: Colonia Primavera. Woolston
SI).
30. Justicia ramulosa (Morong) C. Ezcurra, Bol.
Soc. Argent. Bot. 25: 350. 1988 (Octubre). Be-
— ramulosa Morong. Ann. New York
Acad. 7: 194. 1893. Justicia ramulosa
— A. W. Graham, Kew Bull. 43: 604.
1988 (Noviembre). TIPO: Paraguay. Asunción:
— 706 peor NY isotipos, BML El.
, MO!, NYL KL Û
Beloperone tetramerioides Lindau, Bull. Herb. Boissier 3:
Ezcurra 265
Justicia en Sudamérica Austral
488. "E — е AN Andau) V.
Graham, ull. 603. TIPO: EN а.
Yapacani, roe m, uy о — (holotipo, B
destr., fot. F 89487; isotipo, NY!).
Beloperone = Rusby, Mem. Torrey Bot. Club
6: 103. ›. Justicia coc habumbensis (Rusby) V. A.
. TIPO: Bolivia.
е p E!, K!,
W. 58 Bull. 43: €
15 (holotipo, NY
US!.
Be lopero rone — — “inci w 1159.
27. Justicia Pow tue Mildbr. . W. Gra-
ham, Kew Bull. : r^ un ioo Stein-
7137 bis — B destr., fot. F 8938!: iso-
tipos, BM!, K!, NY!, UC fot. US!).
^
Hierba sufruticosa con rizoma leñoso en la base
y tallos erectos, de 0.40—1 m alt., subcilíndricos.
cuadrisulcados, con dos líneas pilosas longitudi-
nales en su juventud. Hojas con pecíolo de 1-3 cm
long. y lámina elíptica de 4-11 em long. por 2-5
em lat.. aguda en el ápice, cuneada en la base у
algo decurrente sobre el pecíolo, adpreso-pubérula,
con las venas principales prominentes en el envés.
Inflorescencias en espigas apicales y en las axilas
de las hojas superiores, densas, formadas por flores
sésiles con dos bractéolas en las axilas de brácteas
decusadas e imbricadas. Brácteas elípticas a an-
gostamente obovadas, de 1—1.5 em long. por 3-5
mm lat., foliáceas, agudas y mucronadas, pubéru-
las. ciliadas, con las venas principales muy pro-
minentes en el envés; bractéolas lineales de 6-8
mm long., agudas, ciliadas, con la vena media pro-
minente. Cáliz profundamente 4-partido, con los
segmentos lanceolados, de 6-7 mm long.. ciliados.
Corola de color rojo-vinoso a violácea. de 3 cm
long.. con el tubo angosto, de 2 em long. por 2-3
mm diám.: labios de 1 em long., el superior angos-
o. de 4 mm lat., levemente bidentado, el inferior
trilobado, los lóbulos de 4 mm long. por З mm lat.
Estambres con filamentos de cerca de 1 em long..
as anteras con tecas superpuestas, la superior de
1.5 mm long., la inferior más pequeña y espolonada
en la base. Cápsula de contorno angostamente obo-
vado, de menos de | em long., cortamente estipi-
tada en la base, de paredes finas. Semillas apla-
nadas. de menos de 2 mm diám., pardas, con pelos
cortos y gruesos en toda su superficie; retináculos
agudos.
Ilustraciones. Ezcurra, 1993a: ЗАТ.
Distribución, hábitat y fenología. Perú, Bolivia,
sudeste de Brasil. Paraguay y norte de la Argentina.
en ا н htimedos a menos de 500 m de eleva-
. Florece todo el año, principalmente en otoño.
de. mayo a Junio.
Justicia ramulosa se caracteriza por sus flores
rojo-moradas en espigas densas con brácteas im-
bricadas, elípticas. Es una especie variable que po-
266 Annals of the
Missouri Botanical Garden
Figura 12. Justicia polygaloides. —A. Planta con flores. —B. Flor con bráctea, bractéolas, cáliz, corola, estambres
y estilo. —C. Corola abierta con androceo. —D. Cápsula abierta. —E. Semilla. Woolston 799 (SI), Jörgensen 4237 (SI).
Guayculec, Jörgensen 2336 pp (LIL). Jujuy: Ledesma,
Confluencia del Rfo Piedras y Arroyo Pantanoso, Fabris
. А " | 7302 (LP). Misiones: Salto Iguazú, Lillo 10520 (LIL);
mucho a la especie descripta para Guatemala bajo p ы
М І Е пепо Aguirre, Rojas 8070 (LIL). Salta: Orán, Río Pes-
el nombre de Chaetothylax cuspidatus Gibson cado. Pintascavo. Borsini 613 (LIL); Iruya, Pierotti 662(
(Guatemala. Pittier 1792; isotipo, US!) y a Justicia (LIL); San Martín, Gri. Ballivián, El Монет Герпате
teletheca T. F. Daniel (Mexico. Breedlove 56314: 10151 C (LIL). PARAGUAY. Alto Paraná: Reserva Ita-
isotipo, К!), y las tres parecen estar muy estrecha- bó. Caballero Marmori 395 (CTES, MO), Amambay: in
alta planitie et declivibus Sierra de Amambay, in silvis
Punta Pora, Rojas 10463 (B * A K, NY, P, W); Parque
see una distribución amplia en América del Sur,
desde Perú hasta el norte de Argentina. Se asemeja
mente relacionadas.
Material seleccionado. ARGENTINA. Corrientes: Nacional Cerro Corá, Hahn 2481 (G, MO, US). Central:
San Cosme, Paso de la Patria, Costa Toledo, Meyer 9067 V Assomption, dans les bois, Su 2448 (BM, G, К,Р).
(LIL). Formosa: km 139 del FC, Jörgensen 2336bis (SI); Cordillera: ad. marg. laqus Ypacaray, Hassler 3174 (BM,
Volume 89, Number 2
2002
Ezcurra 267
Justicia en Sudamérica Austral
G. К, LIL, MO, NY. P, W): у de Altos, Hassler
2977 (BM. С. К, MO, NY. P, -araguaria centralis, in
regione lacus Ypacaray, Hassler Bos (BM, К, LIL, MO.
¿SL US). Guairá: Tororo, Propiedad de Raul Alvaren-
ga. Soria 2670 (МО). Paraguarí: Guarapi, dans les bois.
Holt 3290 (BM. С. Р): Macizo Acahay, W side of E
peak. Zardini 5882 (MO. US): Mbatoví. Montain Forest
cks of cerrado type, Zardini y Menem: h 3145 (MO):
„| km aro |
San Pe ш Colonia — ra.
Zardin
area.
Parque Nac ш, н
: .U Sy
Woolston 688 x sal SI, US). S
Velásquez 12954 (MO).
de ep.: s . loe.,
31. Justicia riojana Lindau. Bot. Jahrb. 19(48
19. 1894. TIPO:
de Famatina, Los Berros,
derlein 588 (isosintipo. CORD!): entre La En-
crucijada у Las Cuevas, 3 May 1879, Hierony-
CORD!):
Sierra de Velasco, Cuesta de la Puerta de Pie-
dra. 8-11 Feb. 1879, Hieronymus y Niederlein
46 (isosintipo, CORD!).
Argentina. La Rioja: Sierra
Hieronymus y Nie-
mus y Niederlein 546 (isosintipo,
m alt,
erectas, cilíndricas. las jóvenes verdosas y octocos-
Arbusto ramoso, de 1-1.50 con ramas
tadas. con estrías marcadas entre las costillas, ge-
neralmente glabras. Hojas cortamente pecioladas,
angostamente lanceoladas a lineares. de 2—4(—6)
em long. por 0.1-0.3 em lat., agudas, cuneadas en
la base, generalmente glabras, pubérulas solamente
sobre la vena media en el envés. Flores sésiles o
cortamente pediceladas, con dos bractéolas en la
base. solitarias en las axilas de las hojas superiores
de las ramas principales, o de ramitas laterales,
formando en conjunto espigas laxas y foliosas: brac-
j
téolas lanceoladas de cerca de | em long. Cáliz
acampanado, de ca. | em long., levemente 5-par-
tido, con los lóbulos triangulares. subiguales, de З
mm long.. foliáceos. generalmente glabros. Corola
azul violáceo o lilacina de 222.5 em long., pubé-
rula, con tubo corto de cerca de | em long. .. ensan-
chado en una garganta de aproximadamente 0.8 em
diám.. los labios de 1—1.3 em long.. el anterior tri-
la mitad. con la fauce.
lobado hasta convexa y
transversalmente rugoso-venosa. Estambres inser-
tos en la base de la garganta. las anteras con las
tecas subparalelas. una por encima de la otra. la
inferior. generalmente apendiculada en la base.
Cápsulas obovoides, robustas, glabras, de 1.3-1.8
em long. por 5-6 mm diám..
sólida y lateralmente comprimida. Semillas gene-
ralmente 4, suborbiculares. lisas. de 4-5.5 mm
diám.. con un engrosamiento en el borde interno:
retináculos obtusos.
Ilustraciones. — Ariza-Espinar, 1971.
Distribución, hábitat y fenología. | Éndémica de
con la mitad inferior
a provincia de La Rioja. Su distribución está res-
tringida a las sierras de Famatina y Velasco, entre
los 1300-3300 m s.m.
fondos de riachos secos.
Habita en barrancos y e
Florece en verano. de no-
viembre a febrero.
” (Ariza. 1971).
Justicia riojana es muy afin a Justicia tweediana
Vombre vulgar. “Pichanilla
en las flores axilares solitarias y las semillas lisas
con reborde engrosado, pero se diferencia por las
hojas muy angostas y los segmentos del cáliz sol-
dados cerca de dos tercios de su longitud. Es con-
siderada forrajera.
Vaterial seleccionado. ARGENTINA. La Rioja: Fa-
matina, Los Corrales, Cabrera 27228 (Sl): Famatina. La
Batea. Sa. Famatina, Иља 1. 1869 (SI); Felipe
Sierra de Sañogasta, falda W, subiendo desde Aicuña ha-
i as ther, А. 22854, 22863 (CORD); Sa. Fa-
Varela,
matina, camino a * Mina La Mexicana, Kiesling 6402
(SI): Chilecito, Cuesta de Miranda, Cabrera 27112 (S1):
Gral. Lavalle de Cuesta de Miranda. Kiesling 6780
(SI
32. Justicia rusbyi (Lindau) V. A. W. Graham.
Kew Bull. 43: 605. 1988. Chaetochlamys rusb-
yi Lindau, Bull. Herb. Boiss. 3: 1895.
TIPO: Bolivia. La Paz: Guanai, Rusby 1117
(isosintipo, BM!, K!, NY): Santa Cruz, 380 m.
June 1892, Kuntze s.n. (isosintipo, NY! USD).
New York Acad, Sel. ds
ax gii Morong, Ann.
€ iul.
. 1893, non Justicia lanceolata (Cha
TER TIP О; Paraguay. Paraguarí: — Райи»
һе we n Pirayú and Jaguarón, 8 Abr. 1989,
s!)
Morong
бөг (holotipo. NY! isotipo.
Be rene ce adas Bull. S Boiss. 6,
1: 30. 18 TIPO: Paraguay. Cordillera: in —
pr. ( bn de Altos, Hassler 19.36 (lec жеш aqui
designado. G!: E еза, К!, NY!
m
Hierba sufruticosa, de 0.70—1.50 m alt.. con ta-
llos erectos, subcilfndricos, cuadrisulcados. engro-
sados en los nudos, de color verde amarillento, pu-
hérulos. glabrescentes. Hojas con pecíolo de 0.5-3
em long. y lámina ovada de 5-12 cm long. por 2—
6 em lat., aguda en el ápice y obtusa a redondeada
en la base. algo decurrente sobre el pecíolo, verde
clara, pubérula, con cistolitos densamente espar-
cidos en su superficie, las venas principales pro-
minentes en el envés, amarillentas. Flores sésiles.
en espigas truncadas reducidas a flores solitarias
acompañadas de varias brácteas formando un fas-
cículo, opuestas en las axilas de las hojas superio-
res, o en espigas laxas con entrenudos largos, ter-
minales y axilares, con brácteas foliosas muy
similares a las hojas pero menores: bractéolas lan-
ceoladas, de hasta 15 mm long., pubérulas y cor-
tamente ciliadas, con el nervio medio prominente.
Cáliz profundamente 5-partido. con los segmentos
o
Annals of the
Missouri Botanical Garden
lanceolados, rígidos, de 15-20 mm long. por 2.5-
3.5 mm lat., acuminados y densa y cortamente ci-
liados, con el nervio medio prominente. Corola li-
acina de 3—4.5 cm long., con el tubo blanco, de
1.5-2 em long. por 3—4 mm diám., la garganta dor-
1.5-2 cm long..
salmente gibosa, de poco ensan-
chada, y los labios de cerca de 1.5 cm long., el
posterior erecto, de 6 mm lat., muy levemente bi-
lobado, el anterior lila y profundamente trilobado,
reticuloado-venoso en la fauce, con los lóbulos pa-
tentes, em long. por 6-8 mm lat. Estambres
con filamentos de ca. 15 mm long., y anteras con
separadas por el conectivo,
de
las lecas superpuestas y
ambas de 2.5 mm long.,
glabro con estigma levemente engrosado y biloba-
basalmente agudas. Estilo
do; ovario glabro y disco en forma de copa, ondu-
lado.
diám. con el tercio inferior sólida y lateralmente
Cápsulas de cerca de 1.3 cm long. y 5 mm
comprimido, y la porción superior engrosada, ovoi-
de y 4-seminada, pubérula. Semillas subesféricas,
de 3 mm diám., castañas, lisas; retináculos de 3
mm long.
Ilustraciones. Figura 13.
Distribución, hábitat y fenología. Sur de Perú,
sudeste de Bolivia, sudoeste de Brasil y este de
Paraguay. Habita en el sotobosque y en matorrales
de bordes de selva a 200-800 m de elevación, y
es frecuente en claros de selvas degradadas. Flo-
ece principalmente en verano, de noviembre a
abril.
Justicia rusbyi se caracteriza por las flores gran-
des, lilacinas y blancas con la garganta gibosa, fre-
cuentemente solitarias en las axilas de las hojas. A
veces se encuentran ejemplares de esta especie de-
terminados erróneamente como Ruellia en los her-
barios, e incluso fue descripta por primera vez bajo
el nombre de Ruellia lanceolata Morong, tal vez por
el color lilacino y el tamaño grande de sus flores
que se asemejan superficialmente a algunas espe-
cies de Ruellia de la región. El epíteto lanceolata
no se puede utilizar para esta especie en Justicia
por existir un homónimo anterior (Ezcurra, 1993c).
Material se оо PARAGUAY. Alto диче in
Alto Paraná, Fiebrig 5934 (G, US, SI).
arque Nac ional Cerro С ога, Сегго
indez Casas у Molero 6113b (MO, NY); in cam-
po in regione cursus superioris — уз Hassler —
(G); Pe iino те C — го, Sch warez 62 (LIL).
E 2 km E de ( y pois d 2
(CTE ; Ruta Я km (NO)
=
—
| ML).
Central: Nemby, Vavrek 605 (MO, US); Ся pow rio
Industrial, — 3748 (CTES): Villa Elisa, Pedersen
3132 (K. . S1). Concepción: Zw. Rio Apa and Río
editus — 1433 (BM, К, G, К, P). Cordillera:
Cordillera de Altos, Fiebrig 567 (E, G. К, W); Emboscada,
Segunda Compania, па Trompo, Bordas y Schmeda 4093
(CTE Y, US); in regione lacus Ypacaray, Ha: ssler 12458
(BM. 1 UR Cc, К, MO, NY, US); deb vd centrali is, in silva
pr. San Bernardino, — 3017 (B е 7): dads Cordillera
de Altos, Hassler 3740 (BM. К, MO. . Guaira:
Cerro Mumuy, Soria 2991 (G, MO); € olonia Inde penden-
cia, Villarica, Pedersen 10148 (K, MO).
Ihú, cerca ге — Arbo et al. ж (СТЕ, US);
Cerro Palacios, 5 km of Paraguarí, forest on SW slope,
e d A. (MO, US); deem Fiebrig 888
. С. К); Parque Nacional Ybycu’f, gallery forest along
^i rroyo Mina, 3 km N of administration area, Zardini y
Agua yo 11981 (MO, US); Yaguarón, Cerro Corá, Krapo-
vickas 12328 (CTES, US). Presidente Hayes: Paraguay
expedition Capt. T. J. Page, Palmer s.n. (US). San Pedro:
in regione cursus sup. fluminis Jejui Guazú, Hassler 5719
MO, NY, P); Primavera, Alto Paraguay, Wools-
ton 451 (K, LIL, NY, o 1471 (К). S
monte Mhune III, Jörgensen 4303 (L
—
z
‚©
—
dep.: Común cerca
SI, U
IL, MO, NY S).
33. Justicia saltensis T. Ruiz & De Marco, Lilloa
35(2) 13. 1980 (“1979”). TIPO: Argentina.
Salta: Gral. San Martín, Piquirenda a Quebra-
da de Yacuí, 600 m s.m., 3 Feb. 1925, Schrei-
ter 3509 (holotipo, LIL 48919!).
Hierba sufruticosa en la base, erecta, de 0.4-1
m alt., con ramitas tenues, subcilíndricas, levemen-
te cuadrisulcadas, pilosas, glabrescentes a la ma-
durez. Hojas con pecíolo de 0.5-3 cm long. y lám-
ina ovada, de 3—10 cm long. por 1.5—4.5 cm lat.,
cuneada a redondeada en la base, aguda y acumi-
nada en el ápice, pubérula en el haz, y esparci-
damente pilosa a pubescente en el envés. Inflores-
cencias axilares, en espigas unilaterales laxas o en
racimos de espigas, más largas que las hojas, las
flores sésiles en las axilas de brácteas lineales, con
dos bractéolas basales, el raquis levemente hirsuto
o glanduloso-pubescente. Brácteas lineales, de 3—
5 mm, glandulosas y pubescentes, y bractéolas de
2—3 mm long., glanduloso-pubescentes. Cáliz pro-
fundamente 4-partido, con los 4 segmentos angos-
tos, lineales, subulados, de 5-7 mm long.. glan-
duloso-pubérulos. Corola rosada a morada, de l-
1.3 em long., el tubo basal de 4 mm long., ancho,
la garganta de 4 mm long. y 3 mm diám., y los
labios de 6 mm long., el posterior de З mm lat.,
levemente bidentado, el anterior de 5 mm lat., pro-
fundamente trilobado y reticulado-venoso en la fau-
ce, con los lóbulos de cerca de 2 mm long. y lat.
Estambres con filamentos de 5 mm long., las tecas
a distinta altura, la superior de 1.5 mm long., la
inferior oblicua, separada por el conectivo, algo
mayor y apendiculada en la base. Cápsula glan-
duloso-pubérula de 1.5 mm long. por 4 mm diám.,
con el tercio inferior sólido y lateralmente compri-
mido, y la porción superior engrosada, cilíndrica y
i mm
4-seminada. Semillas comprimidas, de 2.5
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
269
Figura 13. — rusbyi. —
con — —JD.
ápsula — —E.
Sem
. Planta con flores. —B. Nudo con bráctea. bractéolas. cáliz y estilo.
illa. Pedersen 3132 (SD. Woolston 451 (SD.
—C.
Gorola
270
Annals of the
Missouri Botanical Garden
diám., pardo-oscuras, densamente hirsuto-pubes-
centes; retináculos de 2 mm long., obtusos.
Ruiz y De Marco,
Ilustraciones. 1980: 15.
Norte de Ar-
gentina, sudoeste de Bolivia y oeste de Paraguay.
Distribución, hábitat y fenologta.
Habita en el estrato herbáceo de bosques de regio-
nes bajas (500-700 m) del Chaco Serrano de Ar-
gentina, y de cerros y serranías del Chaco boliviano
y paraguayo. Florece en primavera, verano y otoño,
de septiembre a mayo.
Justicia saltensis se caracteriza por las flores me-
dianas de aprox. 1 cm long. frecuentemente dis-
puestas en racimos de espigas secundifloras laxas,
con brácteas pequeñas y glanduloso-pubescentes.
Debido a sus características morfológicas, en este
—
trabajo la ubico en la sección Sarotheca (Nees
Benth. (Tabla 1). Como enunciaron Ruiz y De Mar-
co (1980), presenta afinidad con especies de Amé-
rica del Norte como Justicia pringlei B. L. Robin-
son.
Material dies ipe оиы ARGENTINA. Salta: Yaquias-
mé, a 200 n E camino al pie de la montaña, Galetto
186 (Sl); Me tan, 9. 27 km de Metán camino a Gral.
Giiemes, Morrone pus SI); Anta, 7 km al 5 de las Ví-
boras, Vervoorst 4244 (SI), 8448 (LIL); Orán, Piquirenda
(SI). Tucumán: Burruyacu, El Morado, —
Burruyacu, El Río, Peirano 10453 (LI AU U А,
Chaco: Cerro León, Schinini y oe 17793 (CTES);
Parque Nacional Defensores del Chaco, Cerro León, Mi-
sión Tribu Nueva, Duré y Brunner 428 (CTES, MO, PY).
34. Justicia squarrosa Grisebach, Abh. Ges.
Wiss. Goettingen 19: 226. 1874. Beloperone
squarrosa (Griseb.) Lindau, Bot. Jahrb. 19(48):
21. 1894. TIPO: Argentina. Córdoba: Asco-
chinga, Lorentz (sintipo, GOET fot. US!); Pa-
saje, Lorentz y Hieronymus 301 (probables
ran В destr., fot. F 8947!, G no visto
. F 26579!).
Jacobinia ciliata Nees, in DC., Prodr. 11: 333. — non
Justicia ciliata Jacq.. Ноп. Vindob. 2: 47. 1772-3,
nec Justicia ciliata (Ruiz & Pav.) Pers., Syn. 1: 23.
1805. TIPO: Argentina.
1261 (holotipo, K!).
Santiago del Estero, Tweedie
Hierba sufruticosa ramosa de 15—30 cm alt., con
rizoma lefioso en la base y tallos ascendentes, al-
gunos postrados y radicantes en los nudos, subci-
líndricos, cuadrisuleados, glabros o con una línea
longitudinal pilosa. Hojas cortamente pecioladas,
con la lámina angostamente ovada a lanceolada, de
3-7 em long. por 0.5-2 em lat., aguda o acumi-
nada, redondeada en la base y algo decurrente so-
bre el pecíolo, generalmente glabra. Flores sésiles
dispuestas en espigas muy cortas y densas, con-
densadas en el extremo de las ramitas, cada una
en la axila de una bráctea y provista de dos brac-
téolas basales. Brácteas opuestas, lanceoladas, de
cerca de 2 cm long. por 2 mm lat., acuminadas,
ciliadas, con largas cerdas extendidas (de hasta 1.5
mm long.) sobre los márgenes; bractéolas lineales,
mm lat.,
Cáliz profundamente 5-partido con los
de 2 em long. por 1 largamente cerdoso-
ciliadas.
segmentos lanceolados, de 1 cm long. por 1 mm
—
aL, acuminados e hirsuto-pilosos hacia el ápice.
Corola violácea o lilacina de ca. 2.5 cm long., con
tubo algo más largo que los labios y de 2-3 mm
diám., el labio posterior erecto, angosto, de 1—1.2
bidentado, y el anterior
cm long. por 5 mm lat.,
profundamente trilobado, rugoso-venoso y mancha-
do de blanco en la fauce, con los lóbulos de cerca
de 1 cm long. por 5 mm lat. Estambres con fila-
mentos de 7 mm long. y tecas a distinta altura, la
superior de 2 mm long., la inferior algo más larga,
separada por el conectivo y aguda en la base. Cáp-
sula de 1.5 cm long. por 3—4 mm diám., con la
mitad inferior sólida y lateralmente estrechada y la
superior ovoide. Semillas 4, subglobosas, de 2.5
mm diám., castañas, lisas y lustrosas; retináculos
obtusos, de 2 mm long.
90. Ezcu-
Ilustraciones.
rra, 1993a: 354.
Ariza-Espinar, 1971:
Distribución, hábitat y fenología. | Especie am-
pliamente distribuida en el sudeste de Bolivia, oes-
te de Paraguay, y norte y centro de la Argentina,
característica de la provincia fitogeográfica Cha-
quefia occidental. Habita en lugares secos y —
dados entre vegetación de monte entre los 500 y
1000 m s.m.
a junio.
Florece en verano v otoño, de enero
Justicia squarrosa se caracteriza por sus flores
lilacinas grandes con el tubo angosto dispuestas en
espigas capituliformes densas con brácteas angos-
tas y ciliadas. En estos caracteres parece muy afín
a una especie de América Central, Justicia. isth-
mensis T. F. Daniel ( = Justicia panamensis (Lin-
dau) V. A. W. Graham, Panamá. Chagres Valley.
Pittier s.n.; isosintipo, BM),
ambas deberían estudiarse. Es apetecida por el ga-
y las relaciones entre
nado.
ARGENTINA. Catamarca:
oreno, Brizuela 927 (L : L); Capayán, Capayán,
Material seleccionado.
a Paz, El M
Muller 172 (LIL); Valle Viejo, Cuesta del Portezuelo, Hun-
ziker, А. 15313 (CORD); El Juncal, — 845 (S 1). Cha-
co: Napalpí, Campo Largo, Buratovic $ — (LIL). Cór-
doba: Colón, Ascoc ie Giardelli 3 (SD); Pocho,
Sierra de Poc Ap» falda W al noroeste de С inc anf, Hun-
ziker, A. 9810 (CORD); Ischilin, Sierra de Copac 'abana,
Hunziker, A. ous COH Tulumba, ca. de L. V. Man-
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
271
del desvío a Turcal. Hunziker,
A. 25229 (CORD): Capital, C^. de las Rosas, Pierotti 5131
LIL). Jujuy: San Pedro, San Pedro, Ahumada 4481 (SI);
Santa Bárbara, Pto. Nuevo, Cabrera 26. 314 (S
silla, en las inmediaciones
—
Km de 4
Caro 13618 (CORD): Chamical, — De Soria, 8 Km al
SE de Chamical, Hunziker, A. 1 7 (CORD). Salta: La
5 de Cnel. Mc — p 5838 (SI); Ro-
sario de la Frontera, 2.3 Km de la ruta 34, c — a San
i Cristóbal 46675 (LIL): Metán,
(LIL); Anta, n V. González.
Morello s.n. (LIL). Santiago del Estero: Guasaván, falda
oriental ER la Sa. de Guasayán, de S santa Catalina a La-
Heo — 138 (SI); Dpto. Ojo de Agua, Cachi. Di Lullo
7 (5 . El Salve Фан 2. Legname s.n.
igres, З Km al N de Monte Quemado.
;. Pellegrini, Cerro del Remate. ia
man Tum — Santa Rosa, Qda. De La
Higuerita, — 13958 (SI). Tucumán: Trancas. —
de Astrada, Tapia, Diers 46 d — Trancas, Vipos,
ickas y
49,
idor,
Venturi 2686 (LIL). PARAGUA querón: Campo
Loro, limite con Alto Paraguay, Sc a 866 (CTES, US):
Estación Experimental Filadelfia, Vanni 2577 (CT ES): Pi-
"г
cada entre Teniente Ochoa y Mariscal Estigarribia, а 10
km de Mariscal — en — del bosque, Soria
1308 (CTES. MO ueva Asunci ca. Estancia Co-
pagro. 506 km de m бп, Bernard 20202 (( (G. US); Es-
tancia La Madelón, piste Trans Chaco pk 635, — y
Jadin 3123 (G); Fortín Teniente | iso, ruta Trans
— y Bordas 16471 (CTES, G, US); (
. Garay, línea de Hito, frontera con Bolivia, Charpin
Ramella 21485 (G); Ruta Trans-Chaco, Schinini y Bordas
6375 (CTES).
zeneral Eu ugenio
35. Justicia prg (Nees) V. A. W. Graham,
Kew Bull. 43: 604. 1988. Chaetothylax tocan-
tinus — in Martius, Fl. Bras. 9: 153. 1847.
TIPO: Brasil. Goyaz, Pohl (sintipo, W!;
isosintipo, М: Minas Gerais, Claussen
(sintipo, P no visto: isosintipo, K!).
SM.
Hierba perenne, rizomatosa y leñosa en la base,
con tallos erectos subcilíndricos, a veces decum-
bentes y geniculados, de hasta 50 em long., pu-
bérulos o glabros. Hojas opuestas con pecíolo de
5-20 mm long. y lámina ovada de 4—10 cm long.
por 2.5—5.5 em lat., acuminada en el ápice, cunea-
da y algo decurrente en la base, entera, general-
mente glabra, con la nervadura prominente en el
envés. Inflorescencias formadas por flores sésiles
con dos bractéolas en su base, dispuestas en es-
pigas densas enriquecidas por flores derivadas de
yemas axilares supernumerarias. Espigas termina-
les o en las axilas de las hojas superiores. Brácteas
y bractéolas lanceoladas, rígidas, de 4-6 mm long.,
cilioladas, con el nervio medio prominente en el
dorso. Cáliz de 5-8 mm long., con los 4 segmentos
rígidos, lineales, subulados, con pelos glandulares.
7-2.9 cm long.. exter-
namente pubescente, con tubo angosto de 2 cm
Corola rosado—violácea de 2.
long., ampliado en una garganta de ca. 4 mm diam..
el labio superior erecto, de 7 mm long. por 2 mm
lat., el labio inferior ancho, trilobado, de ca. de 8
mm long., con estrías blancas en la fauce. Estam-
bres con filamentos de aprox. 8 mm long.: teca api-
cal perfecta, de contorno ovado, de 1.5 mm long.,
a otra inserta más abajo en el filamento. semies-
téril, más angosta y basalmente caudada. Cápsula
de contorno obovado, de aprox. 7
mm lat.,
mm long. por 2.5
cortamente sólido-estipitada en la base.
generalmente glabra y de paredes finas. Semillas 4
aplanadas, de 1 mm diám.,
con pelos cortos y glo-
quidiados: retináculos de 1 mm long., obtusos.
Nees, 1847a: 26
Ilustraciones.
—
sub Chaetothy-
lax tocantinus).
Distribución, hábitat y fenología. Este de Bo-
livia, sudoeste y centro de Brasil, y extremo norte
de Paraguay. Habita en bosques estacionalmente
secos. Florece en otofio y primavera, principalmen-
te en abril y mayo.
Justicia tocantina es afin a Justicia goudotii.
pero se caracteriza por sus flores rosado-violáceas
grandes de aprox. 3 cm long. con tubo angosto le-
vemente ampliado en la garganta, dispuestas en in-
florescencias kae iformes densas similares a las de
Justicia goudotii. Es la primera vez que se cita esta
especie para Paraguay y Bolivia.
— seleccionado.
in (Lagerenza).
Palese 2970 (G );
mento) hasta Mes
Chaco: Cerro
PARAGUAY.
p-p- (G): Parque Na ional Defensores del Chace
León, Hahn 1530 (G, MO, PY IVIA. Santa Cruz:
Caballero, 8-9 km east of S a at on road to Pulgina.
Wood 10919 (K); Caballero, ca.
Rio Comarapa, Wood 10923 (K).
2 km above Comarapa by
36. Justicia tweediana (Nees) Grisebach, Abh.
yes. Wiss. Goettingen 19: 225. 1874. Adha-
ida tweediana Nees, in DC., Prodr. 395
1847. Justicia tweediana (Nees) Benth.. in
Bentham & Hooker, Gen. Pl. 2(2): 1409. 1876.
Ecbolium tweedianum (Nees) O. Kuntze. Rev.
Gen. Pl. 1(2): 981. 1891. Poikilacanthus twee-
dianus (Nees) Lindau, Bot. Jahrb.
1893. TIPO: Argentina. Córdoba: Río Segun-
do. Tweedie s.n. (sintipo, K!); Parana (“Para-
ma"), Tweedie s.n. (sintipo, К); Buenos Aires,
Tweedie s.n. (sintipo, K!; isosintipo. BM): San
Luis: Aguadita, Gillies s.n. (sintipo, K!; isosin-
BM!, E!).
tipos,
Adhatoda tweediana Nees var. angustifolia Nees, in DC.
Pr
odr. 11: 395. 1847. Argentina. Parana ("Panama"):
Tweedie s.n. (sintipo, К); Buenos Aires. Tweedie s.n.
272
Annals of the
Missouri Botanical Garden
e K!), Gillies s.n. (sintipo, K!); San Luis, Gil-
lies s.n. (sintipo, K!).
Justicia campestris Griseb., Abh. . Wiss. —
19; 225. 1874, non iis erie ris Nees,
Martius, Fl. Bras. 9: 118. 1847. Justicia —
Lindau, Bot. Jab — kie 1894. Ecbolium lo-
rentzianum (Li ind: Kuntze, Rev. Gen. Pl. 3(:
248. 1898. ПРО: ү усен Córdoba. Las Teak
Lorentz (holotipo, GOET no isolipo,
CORD!, fragmento LI
Justicia diamantina Lindau, Bull. Herb. Boiss. 5 (ser. 2):
371. 1905. TIPO: Argentina. Entre Ríos: Diamante,
* 1903, Malme s.n. (sintipo, B destr.;
no vistos).
т
visto;
—
I
isosintipos
Sufrútice o arbusto ramoso, de 0.50-1.50 m alt.,
con ramas erectas, cilíndricas, las jóvenes verdosas
y octocostadas, con estrías marcadas entre las cos-
tillas, generalmente glabras. Hojas cortamente pe-
cioladas, angostamente ovadas a lanceoladas, de 2—
4(-6) em long. por 0.5-1.5(-2.5) ст lat., agudas,
cuneadas en la base, generalmente glabras, pubé-
rulas solamente sobre la vena media en el envés.
Flores sésiles o cortamente pediceladas, con dos
bractéolas en la base, solitarias en las axilas de las
hojas superiores de las ramas principales, o de ra-
mitas laterales, formando en conjunto espigas laxas
y foliosas; bractéolas lanceoladas de cerca de | ст
long. Cáliz profundamente 5-partido, con los ló-
bulos angostamente ovados a lanceolados, subigua-
les, de 5-8 mm long., foliáceos, generalmente gla-
bros. Corola azul violáceo, lilacina a blanca de 2-
2.5 em long., pubérula, con tubo corto de cerca de
l em long., ensanchado en una garganta de apro-
=
ximadamente 0.8 ст diám., los labios de 1—1.3 сп
long.. el anterior trilobado hasta la mitad, con la
fauce blanca, convexa y transversalmente rugoso-
venosa. Estambres insertos en la base de la gar-
ganta, las anteras con las tecas subparalelas, una
por encima de la otra, la inferior generalmente
apendiculada en la base. Cápsulas obovoides, ro-
bustas, glabras, de 1.3-1.8 em long. por 5-6 mm
diám., con la mitad inferior sólida y lateralmente
comprimida. Semillas generalmente 4, suborbicu-
lares, lisas; retináculos obtusos.
Ilustraciones.
Justicia campestris). Dawson, 1979:
1971: 96 (sub
573 (sub Jus-
ticia campestris). Ezcurra, 1993a: 348.
Ariza-Espinar,
Distribución, hábitat y fenología. Norte, centro
y E de la Argentina, desde Jujuy hasta San Luis,
Córdoba, Santa Fé, Entre Ríos y N de Buenos Ai-
res. Vegeta en regiones áridas y semiáridas entre
el nivel del mar y 2500 m (según la latitud). Es
apetecida como forraje por cabras y ovejas en lu-
gares con escasez de gramíneas. Florece principal-
mente en verano, de septiembre a mayo.
Nombres vulgares. “Boca de conejo,” “palomi-
llo," “а Ша,” “ааа,”
dura” (Burkart, 1943, sub Justicia campestris).
Justicia tweediana se caracteriza por las flores
“quiebrarao,” “escoba
medianas, solitarias, axilares, de color celeste-li-
lacino a blanco, y las semillas aplanadas, lisas, con
engrosamiento perisférico. Tiene un alto valor como
forrajera (Burkart, 1943).
A RG ; s NTINA. Buenos Aires:
1 Nicolás, San Nicolás, barrancas, — — (SI);
E Baradero, barrancas, Burkar K, SI). Ca-
tamarca: Andalgalá, Andalgalá, hae en or 31); Ca-
payán, c camino al dique nivelador € án, Legname 7264
LIL); Santa María, El Sarcito, Reales 1055 5 (LIL); Belén,
Pozo de Piedra as, Sleumer y Vert — e П IL). Chaco:
CH, Pampa del Infierno, Meyer 8570 (LIL); Campo del
Cielo, fn, Schulz 1095 (LIL). b ;apilla del Mon-
te, а нра s.n. (LIL); Punilla, Sierra С s ‚ al N del
С° Unpico, Hunziker, А. 8947 (CORD); S n Justo entre
Concepción del Tío y El Tio, Hunziker, A. 10068 CORD);
Totoral, alrededores del agua en el arroyo Macha, 5 Km
W Hunziker, A. 21843 (CORD); Río
Cuarto, entre Berrotarán y Santa Isabel, Ragonese y Pic-
Sur sila seleccionado.
—
de Las Pefias,
> Р.
— 9408 (SD; Tercero Arriba, Río Tercero, Burkart
3 (SI). Entre Ríos: Paraná, Barr s de Paraná,
rior 20647 (Sl); Victoria, Vic toria, Мезе er r 10 2198 (LIL);
1
(LP); Tumbaya, Laguna de Volcán, Kiesling 4345, 5823
(SI). La Rioja: Chilecito, cuesta de Mi K
540, — 4. 22674
iren Da 10 (SD. Salta: Camino a Cachi, pasando Es-
"agnum 9005 (LIL); Candelaria, Unquillo, Schrei-
ler "935 (LIL); Rosario de Lerma, Campo — Venturi
5060 (51). San Juan: 0 Juan, Pa p (LIL).
Fé 3813 (SD. San Luis:
1 auis, Covas 1102 (L AL ); Pe-
=
-=
=
5
AA
— —
оо
р;
کب
derson 1811 (CORD): Dpto. Pringles,
Volcán, Kiesling 4718 (SI); Chacabuco, s arán, 605 m,
— 495 (LIL). at del F
a 16, Legname 7521 (LIL); Copo, Pampa de los Gua-
nacos, Malvárez 675 (LIL). на 1 ntr
fiernillo у Amaicha del », Legname y Cuezzo 5514
(LIL); Tafí, Sa. Del Cajón, Venturi 4438 (SI).
37. Justicia xylosteoides Grisebach, Abh. Ges.
Wiss. Goettingen 19: 225. 1874. Ecbolium
xylosteoides (Griseb.) O. Kuntze, Rev. Gen. Pl.
3(2): 248. 1898. TIPO: Argentina. Santiago del
Estero: “antes que se pasa el Saladillo”, 4/5
Dic. 1871, Lorentz 12 (holotipo, GOET no vis-
to; isotipo CORD!).
-
Beloperone scorpioides Nees, in DC., Prodr. 11: 422. 1847.
Justicia 1 scorpioides (Nees) Griseb., Ae Ges. Wiss.
Goettingen 24: 262. #45 — — scorpioides
L.. Sp. Pl. ed. 2: 21. 1762-63. TIPO: Argentina.
“Tucumán”, Saladillo, — pe ы K!,
fot. US!).
Arbusto de 1-2 m alt.,
mas erectas, 4-sulcadas en su juventud, general-
robusto, ramoso, con ra-
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
mente glabras: ramitas laterales con entrenudos
cortos en la base, alargados superiormente. Hojas
subsésiles o muy cortamente pecioladas, con lám-
ina angostamente elfptica, ovada, obovada u oblon-
ga de 2-4(—6) ст long. por 0.7-1.5 em lat.. glabra
o muy levemente pubérula, glanduloso punteada,
obtusa a redondeada en el ápice, cuneada en la
base. Espigas sésiles, densas, terminales o axilares,
a veces agrupadas en racimos, con brácteas lan-
ceoladas a ovadas, opuestas y decusadas. a veces
alternas por desplazamiento y concaulescencia.
siempre una fértil y la otra estéril de cada par, las
fértiles mayores, de hasta 8(-10) mm long. por 2—
l mm lat., las estériles menores, de 2 mm long. y
| mm lat., el raquis arqueado: bractéolas triangu-
lares, de 2-3 mm long. y 1 mm lat., escabroso-
pubérulas. Cáliz de 3-4 mm long., con los 5 seg-
mentos ovados. agudos, pubérulos. Corola roja. de
34.5 cm long.. con tubo de 2-2.5 cm long.. sub-
cilíndrico. poco ensanchado en la fauce, y labios
de 1-1.5 em long., el posterior apenas bidentado,
el anterior profundamente trilobado. Estambres con
filamentos de 1.3-1.7 em long., y anteras con las
tecas a distinta altura, la inferior oblicua y apen-
diculada en la base. Cápsula de 1.3 ст long. por
0.5 cm diám., sólida y lateralmente estrechada en
la base, superiormente inflada y 4-seminada. Se-
3-4
tuberculado-rugosas. más o menos oscuras.
Ariza-Espinar, 1971: 81.
millas suborbiculares, aplanadas. de mm
diám..
Ilustraciones.
Distribución. hábitat y fenología. Regiones ári-
das del sur de Bolivia, noroeste de Paraguay y no-
roeste de Argentina. Habita en claros de montes
xerófilos bajos y en matorrales áridos entre 700 y
2300 m s.m. Florece principalmente en verano, de
septiembre a mayo
Tal como lo señaló Lindau (1894). Justicia xylos-
tevides es una especie de afinidades poco claras.
tal vez por sus modificaciones en adaptación a la
aridez. Esto la hace difícil de ubicar en el sistema
de Graham (1988). Por sus espigas simples con
brácteas generalmente ovadas, flores grandes у se-
millas tuberculadas se ubica provisoriamente en la
sección Orthotactus (Tabla 1). Es una especie Па-
mativa por sus grandes flores rojas. ornitófilas, que
merecería cultivarse en climas áridos. Es apetecida
por el ganado.
— seleccionado. ARGENTINA, Catamarca:
La Paz, Ramblones, Brizuela 456 (LIL); Fray M. n
Polos ruta 62, Hunziker, A. 22457 (CORD): C apital, ( ia-
amar Vicora s.n. (SI 18623); Santa María. El Sarcito,
Re bai 1043 (LIL); Pachín, entre € catamarca y Cuesta del
Córdoba:
en el camino de Villa
Bañado de Paja y El Chocolate.
Dolores a Chanc ‘anî, — А. 14694 (CORD): Sobre-
monte, 6/7 Km : oeste de la plaza de San Francisco del
Chafiar, rumbo 9 ansilla, Hunziker, A. y Subils
24951 ( CORD): C ruz z del Eje, Camino a Soto, Nicora s.n.
(SI 17632). La Rioja: Chamical, КРА, 5 Km al N de la
Salina La Antigua, Biurrun 584 (CORD): Capital, Dique
Los Sauces, falda de la Sierra de Velasco, Hunziker, A.
1708 (LIL); Gral. Belgrano, Sierra de Los Llanos, ns
Olta y El ое Hunziker, А. 13915 (CORD): Gral. La-
iranda, Kiesling 6783 (SI). Salta: La
Vina, 5 km als de Coronel Moldes, Kiesling 5844 (%1):
. Lomas altas del Cerro Santa Victoria, Legna-
LIL): € tachi, Bajada de Tin Tin, Cabrera 30759
(SI); Cafayate, camino de Santa María a Cafayate, a 2 Km
de Cafayate, Legname 8893 (LIL): Metán, entre Gas del
Estado y Río Juramento, Meyer s.n. (LIL) Molinos, Se-
clantás a Brealito, Meyer 12195 (LIL). Santiago del Es-
suasayán, Sa. de ( suasayán, Avesling 1262 (Sl): Pe-
Sol de Mayo a Los Baños, Peirano 9021 (LIL):
as Termas, Meyer 12789 (LIL); Choy
valle, Cues a de M
-
=
A.
уч
MC
©
-
~
lero:
llegrini,
Rio Hondo, | ба, desde
Puerta Chiquita, yendo hacia Alto Bello, Subils 5666
(CORD). Tucumán: Tafí. Amaicha del Valle, Burkart
5328 (SI): Trancas, Tapia, Rodriguez 597 (SI). PARA-
GUAY. Boquerón: Colonia Menno, Estancia Fe 2 68 km
| 5 і
2: 208 ic S) Made lon; ruta Transe ‘haco, o
E ешп. Û ki N g the — por la picada Boliv-
‚ Vavrek 334 (MO, US).
38. Justicia yhuensis Lindau, Bull. Herb. Boiss.
Ser. 2, 4: 411. 1907. TIPO: Paraguay. In re-
gione fluminis Yhú, Oct. 1905, Hassler 9566
G!; isotipos, BM!, K!, NY! Р!).
(holotipo,
Justicia hylobates Leonard, Sellowia 9: 83, fig. 2. 1958.
TIPO: Brasil. Santa Catarina: prios Popi, pelo
caminho a Sant Antonio, mata, 200-35( p Feb.
1957. Smith, Klein y cia 2» 7 (holo-
lipo. US! isotipo, NY!).
Hierba débil enraizadora en la base, de hasta 60
em alt., con tallos erectos, subtetrágonos, ramifi-
cados. longitudinalmente surcados, pubérulos a
elabros. Hojas cortamente pecioladas o subsésiles,
con la lámina angostamente ovada, de 4—7 cm long.
por 1.2-2.5 ст lat.,
redondeada a levemente cordada en la base, algo
largamente aguda en el ápice.
discolor, glabra o muy laxamente pilosa. Inflores-
cencias en espigas terminales y laterales de hasta
12 em long., laxas, delicadas, con pedúnculos de
2 em long.: brácteas angostamente triangulares, de
2-3 mm long. y 0.5 mm lat., agudas, aquilladas,
elabras o pubérulas; bractéolas de la misma lon-
Cáliz profundamente 4-partido, con los seg-
. por 0.5 mm lat.,
оиа. €
mentos lanceolados, de 3 mm long
agudos. glabros. Corola blanca de ca. 9 mm long.
con tubo obcónico de 3—4 mm long. y labios de 4—
5 mm long.. el posterior triangular, de 2 mm lat. en
274
Annals of the
Missouri Botanical Garden
la base, obtuso; el anterior trilobado, con los ló-
bulos de 1.5-2 mm long. y 2.5 mm lat. (los laterales
algo menores), redondeados. Estambres exertos
hasta la mitad del labio posterior, con las tecas su-
perpuestas, de ca. 1 mm long., separadas por un
conectivo ancho, la superior mayor y oblicua, am-
bas agudas en la base. Cápsula delicada, angosta-
mente claviforme, de 7-8 mm long. por 2 mm
diám., sólida y lateralmente estrechada en el tercio
inferior, glabra. Semillas aplandas, de 1 mm diám.,
con papilas cortas; retináculos de menos de 1 mm
long.
Пизїгасї iones. Wasshausen y Smith, 1969: 96
(fig.
Sur de Brasil,
Habita
Distribución, hábitat y fenología.
noreste de Argentina y Paraguay oriental.
en bordes de selvas de regiones bajas y florece en
primavera y verano, de octubre a mayo.
Justicia yhuensis se caracteriza por sus hojas lan-
ceoladas y sus espigas simples de flores medianas
a pequefias. El nombre Justicia hylobates se con-
sidera sinónimo de esta especie.
Material se — ARGENTINA. Misiones: Ori-
. Deginani 1135 is Guaraní, RP 2
inci | de El Soberbio ca-
ming al Parque Prov. Moconá, ion ie (51); San Ja-
i — Schwarz 4310 (LIL). PARAGUAY. Amam-
ra de Amambay, ad marginem silvarum ad fl.
М. Быр, Rojas 9718 (G); Zw Río Apa und Río Agu
dabán, Caballero Cué, Fiebrig 4 481. 3 (BM, E, G, K). Cor-
dillera: inter rupas in dumetis Cordillera de Altos, Hass-
ler 2135 (G). Cone реш Río Aquidabán, Paso
Horqueta, Palacios 1873 (MO)
E
Especies DUDOSAS O EXCLUIDAS
Justicia flexuosa (Nees) Wasshausen & L. B. Smith (citada
por Wasshausen & Smith, 1969, para Santa Catarina,
Brasil).
Justicia umbrosa (Nees) Lindau (citada por Wasshausen &
Smith, 1969, para Santa Catarina, Brasil).
Estos dos nombres son sinónimos de Poikila-
canthus glandulosus (Nees) Ariza (Ariza-Espinar.
1983; Ezcurra, 1999). Poikilacanthus Lindau es un
género superficialmente muy similar a Justicia, con
polen con características algo diferentes (Daniel,
1991,
pecies de Justicia del Nuevo Mundo a nivel molec-
ular (Mc Dade et al., 2000), por lo que eventual-
mente podría ser tratado dentro de Justicia s.
1998), pero muy relacionado con algunas es-
خر
al.
Pero hasta tanto se estudien las relaciones morfo-
lógicas y moleculares de un número mayor de es-
pecies de Justicia del Nuevo Mundo prefiero man-
tener a Poikilacanthus como género separado de
Justicia. A continuación se da la sinonimia com-
pleta de esta especie.
Poikilacanthus glandulosus (Nees) Ariza, Kur-
tziana 17: 157-161. 1984. Orthotactus glan-
dulosus Nees, in Martius, Fl. Bras. 9: 132.
1847. Adhatoda umbrosa Nees, in DC., Prodr.
11: 406. 1847, non Adhatoda glandulosa
Nees, 1847. Ecbolium umbrosum (Nees)
Kuntze, Rev. Gen. Р]. 1(2): 981. 1891. Justicia
umbrosa (Nees) Lindau, Bot. Jahrb. Syst.
19(48): 20. 1894, non Benth. 1841. TIPO:
Brasil. Sellow 164 (sintipo, В destr., fot. F
8877!).
Nov. Act. Nat. Cur. 11:
Justicia — Nees & Mar
55. 182: 1 Jap:
), non Justicia ed gs Thunb.,
22. 1784 ЕЕ — a — | Mar rt.)
P р Mar Fl. Bras. 9: 136. . TIPO: Bra-
Шош s.n. ‘(holotipo B de 'slr.; — K!).
Maid eun Nees, in وا 9: 148.
17. Ecbolium flexuosum (Nees) ( intze, Rev.
: 980. 1891. Poli anthus Ер (Nees)
. Bot. Jahrb. 18: 57. . Justicia flexuosa
— d & Smith, in R. Reitz, Fl. lustr
Catarinense ACAN: 113. 1969. TIPO: Brasil.
Grande do Sul: e ntre 5 Borin Alegre y Montevideo, “ad
А DM 997 (sintipo, B destr., fot.
i ; Minas ( Villa Rica,
de str.), Martius s.n. (sintipo,
. — Taubaté et Mogy, Riedel s.n.
(sintipo, LE no visto; isosintipo, NY!); Sáo Paulo, ad
. Carlos, Riedel s.n. (sintipo, LE no vis-
—
ripas fl. Urugua
Ft Gerais,
oppidulum S
to).
Ilustraciones. Wasshausen y Smith, 1969: 110
F)
(sub Justicia flexuosa, fig. 15
Literatura Citada
. Las especies de Justicia (Acan-
Argentina. Kurtziana 6: 77—
Ariza-Espinar, L. 197
thaceae) del centro a la
101.
м
19 Notas sobre Acanthaceae ТУ. El género
Paikilas anthus I. Bol. Soc. Argent. Bot. 22: 259-261.
Bailey, 1 949, Manual of Cultivated Plants. Macmil-
lan, т p rk.
Bentham, G. 1876. Acanthaceae. In ho = р & J. D.
— Genera Bison 2(2): 1
Brako, L. & J. L. Zarucchi е К ee Cate of
the Hose Plants and Gymnosperms of Peru
nogr. Syst. Bot. ao А al Garden 45.
Bremekamp, С. B. 1969. An е list of the
Acanthaceae ы ted by › W. М. A. Brooke on her
Ned. FIN Wetensc
trave ^ 10 Bolivia. Proc. Коп. ch..
Ser. C, 72(4): 420—430. [Reprinte n in — Bot. Mus.
VEN к. Univ. Utrecht 331: —430. 1969, |
Burkart, A. 1943. Acantáceas ® como forrajeras
de emergenc iniana 6: 192—202.
~
м
Cabrera, A. L. . Willink. — Biogeograffa de Amé-
rica Latina. Mong 13. O. E. A., Washington, D.C.
Daniel, T. E * ). New and sete Foch Mexican Ac —
hac eae | — Contr. Univ. Michigan Herb. 17:
133-13
1991. A synopsis of Poikilacanthus (Acantha-
me in Mexico. Bull. Torrey Bot. Club 118(4): 451—
458
1993. Mexican Acanthaceae: Diversity and dis-
Volume 89, Number 2
2002
Ezcurra 275
Justicia en Sudamérica Austral
tribution. /n T. P. e R. Bye. A. Lol & J. Fa
(editors), Biological Dive
Distribution. Oxford Univ. |
: Origins and
5, New York & Oxford.
— ‚ Ac ca ‘eae. ; Breedlove, Flora of
Chianas : — 58. Calif. | , San Francisco.
—— 19 qo morphology of Mexican pis
Calif.
Em
^
сеае: — rsity ane systematic significance. Proc.
Acad. — 1. 50(8): 217-256.
00. Additional _ numbers of Amer-
ican в a ‘eae. Sys : 15-25.
————4& T I. Chuang. о Chromosome
New World no 'anthaceae. Ea Bot. 18: 28
B. arfitt & M. A. Baker. 1984. A
their systematic implications in some
| 9: 346—355
numbers of
3-289
— Ts anc ud
North * rican Acanthaceae. Syst. Bot.
| L. Chuang € M. A. Baker. 1990, Chromosome
— rs of — ‘eae. Syst. Bot. 15: 13-25.
Dawson, G. 1965 \canthac ‘eae. In A. L. Cabrera (editor),
Л. Buenos — 5: 312-335. Colece. Ci. INTA, Buenos
Aires.
. 1979. Ас — ‘eae.
Entre Rios 5: 545-
In ^. du (editor), F
. Colece. INTA, Buenos je
De dm N. & T. Ruiz. 1976. Tres nuevas especies del
género Justicia Bow ex L. (Ac — ш рага Argen-
. Esp. Inst. Lillo 1187:
Durkee, L. H. 1986. Acanthaceae. *
Flora болайык, Family 200. Fieldiana, Bot., n.s.
18: 1-87.
— (editor),
Ezcurra. €. 1988. Novedades nomenc е еп especies
de Justicia Goes ie) del 1 de la Argentina.
Bol. Soc. — Bot. 25(3—4): : Sn
1989. Ruellia sanguinea (Acanthaceae) y espe
|
cies — ionadas en Argentina, Uruguay y sur de Bra-
sil. Darwiniana 29: 269-287.
1993a. Acanthaceae. In
la LER ne la de Jujuy 9:
L. Cabrera —
Flora de 278-359. Colecc.
INTA, donde Aires.
. Notas sobre la posición sistemática y dis-
geog ográfi ica — lilloi (Acanthaceae).
Argent. Bot. 29(3—4): 187—190.
1993 © Systematics к Ruellia (Acanthaceae) in
aiher South America. Ann. Missouri Bot. Gard. 80
787
ibus ión
Bol.
x
1-845.
1999, Acanthaceae. In F. O. Zuloaga & O. Mo-
rrone (editors), Catálogo de las aes Vasculares de la
Re RA a Argentina. Monogr. Syst. Bot. Missouri Bot.
Gard. 74: 1-14.
& й sta-Castellanos. 1997. — —
yJ. ae front, Acanthaceae), 1
vas citas para Argentina y bon. y notas E su
posición. sistemática. Bol. Soc. Argent. Bot. 33(1—
107—111.
07—
Gentry, A. H. 1982. Neotropical floristic diversity: Phy-
кее ‘al connections between central and South
America, Pleistocene climatic fluctuations, or an acci-
dent of yere orogeny? Ann. Missouri Bot. Gard. 69:
551—593.
Gibson, D. 912. е
plants Ш. Fein. _ y
Graham, V. A. W. — and infra-generic
classification. of — (Acanthaceae). Kew Bull. 34:
991-624.
In Studies in. American
4.
Grant, W. F. 1955. A cytogenetic study in the Acantha-
ceae. ion 8: 121-149.
Hay, R. & К. A. Beckett. 1975.
Headers Digest Encyclo-
— of Garden Plants and Flowers. The Reader's Di-
est Association, London.
Hite heock, A. S. & M. L. Green. 1930. Standard species
of Linnean genera of Phanerogamae (1753-54). Int. Bot.
Congr. PELA (Е ngland) 1930, Mount Prop. Brit.
Bot.: 110
Holmgren, P Ps N. H. Holmgren & L. C. Barnett. 1990.
Index Herbariorum. Part I: The He of the World.
3th ed. New York Bot. Gard., Bron. New
eón-Yánez. 1999,
antas Vasculares del Ecuador. Monos Syst. Bot. Mis-
York.
Jörgensen, P. M. & S. Le Catálogo de las
Jot. Mi
souri Bot. Gard. 75.
E C. 1995.
Cipó, Minas Gerais,
14: 18 l- ne
| —
Acanthaceae. In Flora da Serra do
Brasil. Bol. Bot. Univ. Sâo Paulo
. 1951-1958. The Ac + of Colombia.
Contr. U. ^y Natl. Herb. 31(1-3): 1-781.
. М. & C. Ezcurra. 1991.
ment as an aid in the study
in Justicia and Ruellia (Acanthac eae). Bot. J. Linn. Soc.
105: 285-288
r nzyme etching treal-
of seed surface sc нея
59
Catálogo de las Acantáceas Argentinas.
: 21-66. 9 bare.
1893. Beitrage zur Systematik. der. Acantha-
. Jahrb. Syst. 18: 36-04.
894. Beitrage zur argentinischen Flora. Bot.
1. aah, Beibl. 48: 8 —
895. Acanthaceae. . Engler &
Die Natûrlic hen st r n 43b): 274—
rantl.
Linné 1753. Species Plantarum, ed. 1. ES niii TrA
уп, жи
otti 976 nueva espec ie de Acanthaceae:
isa
ina
—— е Lotti. Publ. Esp. Inst. Lillo 1187:
336€
"i Dade, a A. 198
— — eid Costa Rica, with taxonomic notes.
Syst. Be n. : 489—497.
& T E Daniel.
. New species of Justicia and Razisea
é 1998. ер the Justicieae
(Acanthac — Amer. J. Bot. 85: —144.
(Sy E. Masta & К. » pe 2000. Phy-
logenetic relationships within the tribe Justicieae
(Acanthaceae): Evidence from molecular
morphology, and cytology. Ann. Missouri Bot.
35—458
sequences,
Gard. 87:
Nees, C. G.
Acanthaceae. /n C. Martius (editor).
Flora Brasiliensis 9 1-164.
47b. Acanthaceae. In A. P. de Candolle (edi-
tor). p m Systematis Naturalis Regni Vegetabilis
Ll:
Parodi, L. M. J. Dimitri. 1980. Las plantas cultiva-
das. hs ic alon edia — de Agricultura y Jardinerfa
| (2). — ed. ACME. Buenos Ai
Peixoto, M. 1998.
pecies de — (Acanthaceae) del noreste
Darwiniana 36(1—1): 13—28.
————, C. Ezcurra & D. Wasshausen. 1998. а Jlo-
FEM (Acanthaceae), nueva combinación y nueva cita
Darwiniana 35(1—4): 15 * 54.
a
3
D iudi morfo-anatómicos en es-
Argentino.
зага la flora argentina.
Pennington, R. T., D. E. Prado € €. Pendry. 2000.
Neotropical seasonally dry forests and Quatemary veg-
elation changes. J. Biogeogr. 27: 261—
Petriella, B. 1968. El polen p» las Ac e 'eae =
nas. Revista Mus. La Plata, Sece. Bot. 11: 51-6€
Piovano, М. & 1. M. 1991.
numbers in Argentinean Acanthaceae. Syst. Bot. 16:
Bernardello. C —
89-97.
Profice, S. R. 1993.
Estudos em Acanthaceae Novas
276
Annals of the
Missouri Botanical Garden
— s e novos sinonimos, Justicia L. Eugeniana
20:
Raj, B. 1961. Pollen — studies in the Acan-
thaceae. Gran na Palynol. 3
‘urther pei ESA to the pollen mor-
phology of the Acanthaceae. J. Palynol. 9: 91-141.
Rambo, 1964. Acanthaceae Riograndenses.
Bot. т 36.
Rizzini, C. T. 1948. Disquisito circa Acanthacearum ali-
quot genera Brasilensis. Arch. Jard. Bot. Rio de Janeiro
95-372.
—
Iheringia
. 1949. Contribuicáo ao conhecimiento da tribo
Justicieae (Acanthaceae). Arch. Jard. Bot. Rio de Ja-
neiro E 37-€
951. Sinopse parcial das Acanthaceae brasilei-
ras. —* ‘nia =i 45-180.
Ruiz, T. & N. De Marco. 1980. Una nueva especie de
Justicia a dedo 'eae) en la Argentina. Lilloa 35(2): 13.
1979-1980
Sell, Y. 1969. Te 's complexes inflorescentiels de quelques
Acanthacées. Etude particulière des phénomènes de
condensation, de race misation, d homogé néisation et de
tronc — s Sei. Nat. Bot. sér. 12, 10: 225—300.
. Tendences e Sale 's parmi les complexes
— sce a ‘ls. Rev. _ n. Bot. 83: 247-267.
›. Acanthaceae. In Lundell (edi-
pecies and new combinations of Jus-
ticia (Acanthaceae) from the Venezuelan Guayana. No-
von 2: 62—80).
& L. B. Smith. 1969. ч ne In R. Reitz, Fl.
Il. Catarinense | (ACAN): 3-134.
& C. Ezcurra. 1997. n. names and new com-
binations in Justicia sects. Simonisia, Plagiacanthus
and Orthotactus a from southern South
rit ш а. Candollea 52: 171—179.
Webb, C. J. 1995. Checklist of dicotyledons,
sperms, and pteridophytes naturalized or casual in New
Zealand: additional records. New Zealand J. Bot. 33:
151-182
gymno-
APÉNDICE |
LISTA DE. ESPECIES
Justicia pi unde (Nees) Lindau
Justicia axillaris (Nees) Lindau
Justicia baenitzii (H. Winkler) С.
Justicia brasiliana Roth
Justicia carnea. Linc
Justicia chacoénsis Wessbausen & C.
Justicia comata (L.) Lamar
Justicia corumbensis (L паш) Wasshausen € C. Ez-
№ –
Ezcurra
> Y
— Л
Ezcurra
^I oen
^
TU
curra
9. Justicia cuspidulata (Nees) Wasshausen
10. Justicia dumetorum Morong
11. Justicia floribunda (С. Koch) Wasshausen
12. Justicia gilliesii Asel Bentham
13. Justicia glaziovii Lin
14. Justicia ш (re me kamp) V. A. W. Graham
15. Justicia goudotii : Graham
16. Justicia hassleri a indan) V A, W Graham
17. Justicia hunzikeri Arizi
20. Justicia laevilinguis (Nees) Lindau
21. Justicia lilloana Ariza Espinar
22. Justicia lilloi (J. L. Lotti) C. Ezcurra
3. Justicia lythroides (Nees) V. A. W. Graham
4. Justicia mandoni (Lindau) Wasshausen & C. Ezcurra
5. Justicia oblonga (Nees) Lindau
6
7
—
. Justicia oranensis De — o & Ruiz
Justicia pectoralis Jac
28. Justicia phyllocalyx (1 DNE Wasshausen & C. Ez-
curra
29. Justicia polygaloides (S. Moore) L —
30. Justicia ramulosa (Morong) €. Ezcu
3l. Justicia riojana Lindau
: Вицо ta rusbyi (1. — V. A. W. ш
z & De Маг
3 A. W. Graham
Зб. Justicia tweediana ( lees) саан
37. Justicia xylosteoides Grisebach
38. Justicia yhuensis Lindau
APENDICE 2
ÍNDICE DE EXSICCATAS
Lista de ejemplares examinados con — de cole "cc io-
nista (en orden alfabético) y número de colec
mero entre paréntesis corresponde a la especie a la que
pertenece cada eje mplar según la numeración en la lista
de especies. Los ejemplares tipo se indican con asterisco
*
Abbiatti y Claps 221 (18). Aguilar 566 (4). Ahumada
4481 (34); 8234 (3). Anderson 1811 (36); 1936 (17); 4044.
12). Arbo y Tur 2214 (20). Arbo et al. 906 (2); 1652 (10);
1657 (20); 2664 (32). Archer y Rojas 4847 (20). Arenas
143 (10); 287 (9); 716 (32); 1097 (10); 1189 (4); 1457
(10); 2283 (27). Arnott s.n.* (20).
Bacigalupo 997 (29); 1376 (20). Bacle s.n.* (20). Baer
80 (26). Balansa 2448 (30); 2449 (20); 2451, 2451a (20);
2456 (16); 2459, 2459a (4); 2468 (7); 3296 (30). Balegno
1191 (37). Bang 1215* (30). Basualdo 1876 (4); 2546
(23); 2658 (16); 3485a (13). Beck 5373 (10). Bernardello
y Galetto 800 (10). Be — 18720 (4); 18891 (4); 19330
(4); 20142 (8); 20202 (34); 20335 (21), 20425 (22); 20457
(8). Bertoni 1168 (20); 1378 (7); 2433 (20); 2513, 4024
(4). Billiet y Jadin 3057 (15); 3123 (34); 3126 (37). Biu-
rrun 584 (37); 1372 (12), Blanchet s.n.* (7); 2575* (1);
2979* (13). Bordas 3748 (32). Bordas y Schmeda 4093
(32). Borsini 613 (30). Botta 138 (34). Brizuela 456 (37);
927 (34); 1132, 1266 (12). Brooke 5586* (19); 5677*
3). Browne s.n.* (7); 2007 (19). Bruch y Carette 24 (12).
eee І т (14); 1534 (4). Buchtien 1409* (3). Bura-
34). Burkart 5328 (37); 8498 (36); 12528 (34):
: J 13368 — 19998 (355 14171 (2); 14280
—
=
27112, 27228 (31): 27954 (26) * ) (37); 31312 (18);
32149 (26); 32659, 34155 (24). Calne та y Sáenz 29211
(4). Carter s.n. (10). Castellanos s.n. (15); (36). Claussen
s.n.* (35). Covas 1102 (36). € ristóbal a (22); 442 (12).
Cuezzo y de la Sota 1527 (23). Chamisso s.n.* (5). Char-
pin y Ramella 21485 (34); 21610 (6); 21733 (22); 21742
(6). Chodat s.n. (20).
de la Sota 96 (18); 1138 (15). De Marco 10654 С (4).
Degen 1032 (4); 1522 (23); 1611 (23); 3280 (6). Degen y
Mereles 3250 (8). Deginani 1135 (38); 1320 (5); 2080 (5).
Descole 3187 (7). Di Lullo 27 (34). Díaz 10191 (33). Diers
46 (34); 225 (21). Duarte 1705* (7). Duarte y Pereira
1927* (15). Duré y Brunner 428 (33).
~
Volume 89, Number 2
2002
Ezcurra
Justicia en Sudamérica Austral
Echegaray s.n.* (12); s.n. (36). Eyerdam y Beetle 22275
(19).
d (302 (30). Fe rnánde az 144 (23). Fernánde Z he asas
4114 (14). Fernández С \ 113b
(32. Fernández Casas et a 7516 (20). Pernice; a (16).
E 4 (10); 5730, 50206 (5); 5929 (25): 5
E (2: " 6081 (7); 6173 (23); 6174 (4). Fortunato 5880
17): Fortunato y Palese 2970 (35). Frenguelli 128 (14).
Galetto 186 (33). Giardelli 1173 (34). Gillies s.n.* (12):
s.n.* (36). Glaziou 13073* (13); 13076* (1); 21870* (28).
го 154 (20): 680 (8); 726 (4).
Hahn 787 (10); 1530 (35): 2481 (30). — 138, 916,
1080%, 1138 (10): 1579, 1593 (1): 1650 (10): 1907 (2):
32): 2135 (38); 2754 (8): 2755* (10); 2
2071 (30): 3016 (4); 3017 (32); 3053, 3057 (16
: 3508 (20); 3740 (32); 3920 (7); 4889 (20): 5
5207 (4): 5719 (32): 5984 (20); 5985 (7); 6476 (10); 6813.
6813a, 7106 (2): 7560. 7624 (20): 7633 (29); 7749 (32):
7834, 8003 (9): 8160, 8160a (20); 8423, 8423a (28); 9271
(2): 9508* (38): 11105 (4); 11701 (80); 11740 (4): 12458
;: 12530 (16). Hayward 1444 (20). Herrera 232 (18).
Hicken 12911 (20). Hieronymus y — in 46* (31):
Hunziker,
; 4721 (12); 8947 pee 9810 (34):
12322 (12); 13915,
15096 (12); 15313. 160647 (34): 2
: 22457 (37); 22074 (36); 22854. een (31):
34). Hunziker. А. y Caro 13560*, 13562 (17):
118 (34). Hunziker, A. у Cocucci 16388 u2) Hunziker.
A. y Maldonado 16245 (17). Hunziker, A. y Subils 24951
(37). Hunziker, J. 950, 959 (11); 10891 (23). Hunziker, J.
y Gamerro 11669 (21).
ludica у Ramadori 354 (3).
Jórgense n 698 (2); 982 (360); 2332bis (10); 2335 (20):
2336 pp (30): 2336 pp. 2337 (B); 2336bis (30): 2338 (15):
2832, 3278, 3796 (7); 4237 pp (20); 4237 pp (29): 4303
—
14792 (34);
(32).
Kerr 108% (10). Kiesling 1262 (37): 3013 (17); 4345.
1718. 5823 (36): 5012 (12); 5502 (18); 5618 (7); 5838
Н 58H (37): 6402 (31): 6410 (30): 67! 30 (31); 6783
12328 (32); 12525
15161 (25); 17978 (25);
‚ Krapovickas y Cristóbal 11812 (7
14290 (37); 46075 (34). — kas et al.
EN 3890 (20); | 1227 00), Kuntze s.n. * (1): s.n.
(16). s.n.* 19), s.n.* (29), s.n.* (32).
de *
1
1 300% (20). — y Cuezzo 5514 (36)
8403 H 1): 8564c* (26):
15372 (22): 5907 (7): 6052. 6857 (15):
726 1 (36): 7 7521 (36); 8893 (37); 9005 (36):
| (30). Lillo 4417 (22): 6082. 7894 (12): 8087*
: 10520 (30). Legname у López 8704 (15). Linden
1633* (7). Lorentz s.n.* - t); 12* (37); 118% (36). Lorentz
y Hieronymus 301% (34). Luschnath s.n.* (5).
Macedo 952* ya [үә (13): 1617* (28). Malme
s.n.* (36); 3029* (8). Malvárez 675 (36). Mandon 297*
1). Maranta 295 у ). Marmol 8734 (19). Мае 7 Cro-
ES 3376 (26). Martius s.n.* (1), s.n.* (4), s.n. * (5), s.n.
(19), s.n.* (39). rig 678 (4). Mathews 3152bis* (9).
л
As
e
eda a
ыл E ч
^^
©
I
rg Vied* Mereles 1198 (4); 1208 (7); 2054
(11): 2430 (20); A (6): 3270 (7); 3904 (4). Mereles у
Degen и y: 5563 (10). Mereles y — 2795 (6).
Meyer s.n. (37); 2004 (7): 2998 (15): 3197 (20):
8570 (36):
8669 (15); 9067 (30); 10198 (36); 10281 (20); 12195,
12789 (37); 13853, 15125* (24); E (20); 16235 (10):
10416 (18); 16492 (34); 21829 30
Montes 841 (25); 7157 AU 9800, ,
16383 (25). Moore ps * (29); 941 * (20); 1047* (6).
161 (20): 605 (7): 1719 (10): 7182 (4): 7190 (7). Morello
s.n. (34); 966 (24); 16936 (20). Morong 200a (4); 435 (20);
667* (32): 706* (30); 1031 (20); 1538* (10). Morrone 695
(38); 3087 (33). Mroginski 2: 4). Múlgura 675, 738
(17): 927 (15); 1039 (20); 1079 (24); 1891 (5). Muller 172
Nicora s.n. (SI 17632), s.n.
(SI 186023) (37); s.n. (SI
Charpin 10344 y
nell 5368 (18). — 1280 (23). Osten y Rojas 8978 (16).
Palmer s.n. (32). Parodi 8511 (10). Pedersen 3132 (32);
5994 (7): 10148 (32). Peirano 9021 (37); 10453 (33). Pie-
rotti s.n. (12); 83 (15); 4204 (4); 5131 o 6020 ~
1172 (22). Pire 3813 (36). Poeppig s.n.* ( ohl s
(23); s.n. (35); 196% (5); 1849* (1); 1989* a ч Porta 39
(23).
Quarín 367 (20). Quarín et al. 1510 (20).
y Piccinini 9408 (36). Re ie 94 k еч
Ragonese
w
أ
=
Y
л
o
~
е,
~
c
g
=.
=
w Y
* —
-——
>
—
—
~~
/;
cd
N
s.n. — 58^ at 1):
lf ©
> hae т 41
2198 (29): 2199 (20); 2
2930 (20): 8070 (30): 97 18 (38):
3): 10403 (30); 10688 (5); 10864
(13): 10929 (1); 2 (22). Romanczuk 188 (2). Ruiz
Huidobro 2128 (4); im 50 (20); 3348 (20); 3714 (7); 4249
(4): 4301 (4); — a Leal 11819 (12); 22103 (21).
Rusby 1117* (32); 14 20).
166* (14). S. E Herb. Martii 1047* (4). Salz-
mann s.n.* (7). Saravia Toledo 2138 (22); 10779 (10).
Schinini 8327 (2); 21794 (2); Aq 23450 (25). Schi-
nini y Bordas 13314 4 ); 14938 (27); 14973
(6); 16375 (34); 16471 (34); 17793 (33); 17791 (0) 18090
(6); 20344 (14): 20534 (9); 20628 (7); 25242 (7). Schinini
y Caballero Marmori 27168 (2). Se — y — ( ros
Schinini y Palacios 25788 (27); 25857
Se cum da 866 (34). Se homburgk : s.n.* (7).
— o Schreiter 1105 (21); 3509*,
4 192 D 9358 (30): 10866 (10); 11228 Gy
2 (26). Schultz 1095 (36); 3235 (8); 7
(10 " 16244 (4). Schulz 5484 (14); 15846 (8);
18213 (8). Schwarz 94 (4); 1639 (4); 4176 (5); 1310 (38):
5087 (23): 5706 (25); 9275 (4); 9394 (4); 11702 (32).
Se hwindt 3065 (25). *ellow s.n. us. s.n.* (4), s.n.* (7),
п. (39), 97% (3): 120*
+ 174* (1); 214%, 252%,
y Vervoorst 2425 (30).
Smith. Kla in y Sehnorrenberger 11727* (38). Solbrig 51
(36). Solomon 7055. 7056 (28); 14633 (8). Soria 280 (14);
1308 (34); 1727 (14); 2242, 2588, 2037 Vi 2070 (90);
2720 (4); (32); 3303 (7); 3310 (32); 3894 (4). *
Zardini y Ortiz 2036 (9). Soriano 828 (12); 845 (34);
2); e (7 i
10152 —*
S. col.
Y
Е
<>
(37122717 (10 jè 27). Stange s.n. (1 i
6973 eS y 7137 hist Agel Stuckert 17193 (17). Subils
3000 (37). — вл (Th
Teague 572 (20). Tressens et al. 5047 $ 1). Troll 1688
(10). Troncoso s.n. vs 60286) (12); | à 58, 27231 (2). Tron-
созо у Es — EIE (2). s 10 m Tweedie s.n.
(4). (11). s.n. (15). (20). s.n.* (20), s:n."
278 Annals of the
Missouri Botanical Garden
(36 * (37); 752 * (4); 1159* (37); 1260 p.p.* (12: — Beloperone squarrosa 270
1261: > К (34): 1262* (15). Beloperone tetramerioides 205
Ulibarri 1509 (15). Beloperone viridissima 238
lanni 2577 (34). Vanni y Ferraro 693 (14). Varela 495 Boca de conejo 272
(36). Wasshausen 1955 (3). Vavrek 334. (37); 605 (32 Brazilian plume 241
Venturi 88 (18); 1150 (22); 1724 (15); 1799 (24); 1814 Cyphisia 234
(12); 2686 (34); 2835 (18); 4438, 5060 (36); 5217 (19) Cyrtanther 234
5 (15); 5591 (7); 5853 (18); 5866 (33); 5868 (34); — magnifica 240
6184, 7080 (15); 7826 (12); 8121 (18); 9338 (24); 9631 Cyrtanthera — var. minor 240
(15). Ve — 4221 (22); 4244, 8448 (33). Vervoorst у Cyrtanthera pohlia 240
Cuezzo, 7725 С (3). Villa С e nzo 1119 (12); 1750 (22). Cyrtanthera — var. obtusior 240
Villa Carenzo y Le е 1133 (37). Cyrtanthera pohliana var. velutina 240
lasshausen 1955 We dde] 3117 (22). West 8537 — Chaetochlamys 234
(4). White 550% (3), S ink 133 (3). Wood 10919 (35); Chaetochlamys lillo 259
10923 (35). Woolston 193, 193C (4); 451 (32); 688 (30); Chaetochlamys marginata 254
770 (7); 799 (29); 1383 (7); 1471 (32). Chaetochlamys rusbyi 267
Y barrola 324 (20); 1891 (7); 2866 (2); 3504 (20). Chaetoc — tucumanensis 259
Zardini 4973 (30); 5882 (30); 11349 (4). Zardini y Chaetothy 234
Aguayo 9806 (32); 10622 (32); 10730 (4); 11981 (32). Ce boliviensis 253
Zardini y Soria 4442 (4). Zardini y Velásquez 9935 (20); Chaetothylax cuspidatus 266
9943 (16); 12094 (4); 12954. (30); 13143 (30); 13803 (15). Chaetothylax hatsbachii 200
Zardini et al. 2592, 2593 (10); 12782 (4). Zuloaga 1944 — Chaetothylax huilensis 253
(7); 2111 (11); 2188 (23); 3230 T 5270 (4). Chaetothylax tocantinus 271
Chaetothylax umbrosus 252
APÉNDICE 3 — vestitus 253
; | PEE Е A Dianth 234
ÍNDICE DE NOMBRES CIENTÍFICOS (NOMBRES VÁLIDOS EN UL — 242
NEGRITA, SINONIMOS EN BASTARDILLA) Y DE NOMBRES Dianihera rraminifolia 257
8
Tu Dianthera laevilinguis 257
Acelica 234 — Dianthera nodosa 239
Adhatoda 234 — Dianthera — 257
Adhatoda flexuosa 274 Dianthera pecto 262
Adhatoda gilliesii 250 Dianthera rice 264
Adhatoda glandulosa 274 Ecbolium 234
Adhatoda tweedianc 27] Echolium comatum 242
Adhatoda twee diana var. angustifolia 271 Ecbolium flexuosum 274
Adhatoda umbro 274 Ecbolium lorentzianum 272
Albahaca de vaca 235.250 Ecbolium pectorale 262
Alfalfillo 258 Echolium tweedianum 271
Alfalfita әтә Ecbolium umbrosum 274
filla 335.272 — Ecbolium xylosteoides 27
Amphiscopic 234 Escoba dura 27:
Amphisc opia ae quilabris 2360 thesia 234
Amphisc ора ciliata 240 Ethesia carnea 240
Amphise ора strobilacea 237 Нешгейа 234
Amphiscopia venosa 237 Heinzelia lythroides 259
Bandera espa añola 235, 249 — Heinzelia ovalis 259, 200
Beloperone 234 — Iehiyuyo 256
Be ын гопе =: 247 — Jacobinia 234
Beloperone amhe 239 — Jacobinia — 236
Be e rone ne did "rstiae var. ciliaris 239 Jacobinia « 240
Beloperone amherstiae var. debilis 239 — Jacobinia ciliata 270
Beloperone amherstiae var. graciliflora 239 Jacobinia fest 239
Beloperone amherstiae var. lanceolata 239 Jacobinia — teata 237
Beloperone baenitzii 238 Jacobinia magnifica 240
Beloperone brasiliana 239 Jacobinia obtusior 240
Beloperone cochabambensis 265 Jacobinia pauciflora 249
Beloperone corumbensis 243 Jacobinia pohliana 240
Beloperone hassleri 254 — Jacobinia tenuistachys 238
Belope rone kerrit 247 Jacobinia velutina 237, 240
Beloperone mandoni 200 Justicia 234.
Pii matthewsit 267 Justicia acuminata 242
Beloperone pseudociliata 205 Justicia aequilabris 236
Beloperóns ramulosa 265 Justicia alopecuroidea 254
Beloperone riparia 241 Justicia allocota 259
Beloperone scorpioides 272 Justicia amambayensis 245
EX spathulata 274 Justicia anagallis 257
Volume 89, Number 2 Ezcurra
2002 Justicia en Sudamérica Austral
Justicia ascendens 257 Justicia riparia
Justicia axillaris 237 Justicia rizzint
Justicia baenitzii 238 Justicia rusbyana 2
Justicia boliviana 245 Justicia rusbyi 207, :
Justicia brasiliana 235, 2309 Justicia saltensis >
Justicia campestris 237,272 Justicia sarmentosa
Justicia carnea 235, 240 Justicia sarotheca
Justicia carthaginenesis 247 Justicia scorpioides
Justicia ciliata 270 Justicia schreiteri
Justicia cochabambensis 2605 Justicia spathulata
Justicia comata 242 Justicia squarrosa
Justicia corumbensis 243. 256 — Justicia strobilacea
Justicia cuspidulata 245 Justicia teletheca
Justicia chacoénsis 241 Justicia tetramertoides
Justicia diamantina 272 Justicia thunbergioides
Justicia dumetorum 247, 261 Justicia tocantin:
Justicia echege i 250 Justicia tucumanensis
Justicia flexuosa 274 usticia tweedia 235, 271, 3
Justicia floribunda 235, 249 Justicia umbrosa 252,
Justicia furcata 247 Justicia velascana
Justicia gilliesii 235, 250 Justicia velutina
Justicia glabribracteata 237 Justicia venusta
Justicia glaziovii 250 usticia venosa
Justicia glutinosa 252 Justicia vestita
Justicia goudotii 252.211 Justicia xylosteoides
Justicia hassleri 254 usticia yhuensis DTP
Justicia humifusa 212 Le ptostach ya 234
Justicia hunzikeri 254 Leptostachya coma 242
Justicia hylobates 273 — Leptostachya martiana 242
Justicia isthmensis 270 Leptostachya martiana var. hispida 242
Justicia jujuyensis 255. 245 Leptostachya martiana var. macrophylla 242
Justicia kuntzei 250 leptostachya parviflora 242
Justicia laevilinguis 257 Libonia floribunda 249
Justicia lanceolata 267 Lophothe cium 234
Justicia leonardii 255 Lophothecium boliviense 256
Justicia lilloana 258 — Lophotecium paniculatum 250
Justicia lilloi 26. Orthotactus 234
Justicia lithospermifolia Orthotactus aequilabris 230
Justicia lorentziana Orthotactus arnottianus 237
Justicia lythroides Orthotactus ciliatus 240
Justicia mandoni Orthotactus lanl 274
Justicia marginata 1 Orthotactus strobilaceus 237
Justicia — 250 Orthotactus venosu. 237
Justicia nemoralis var. tomentosa 256 Palomillo
Justicia nodosa 239 Pichanilla
usticia oblonga 201 Plume p
Justicia obtusifolia 257 Poikilacanthus 203,
Justicia odonellii 200 Poikilacanthus glandulosus
Justicia oranensis 235. 202 Poikilacanthus phyllocalyx
Justicia ovalis 260 Potkilacanthus flexuosus
Justicia ovata 25 Poikilacanthus tweedianus
Justicia paludosa 257 Psacadocalymma k
Justicia panamensis 27 Psacadocalymma comatum 242
Justicia paniculata Psacadocalymma pectorale 202
Justicia paraguayensis Pupilla 234
Justicta pauciflora Quiebrar: 212
Justicia pectoralis Ruellia ee eolata 267
usticia peruvian Rhytiglossa 234
Justicia phyllocalyx Rhytiglossa acuminata 242
usticia platicarp Rhytiglossa anagallis 257
Justicia polygaloides Rhytiglossa axillaris 237
Justicia pringlei Rhytiglossa campestri 272
Justicia pseudociliata Rhytiglossa cuspidulata 245
usticia ra Rhytiglossa hookeriana 247
Justicia reitzii Rhytiglossa laevilinguis 257
Justicia repens Rhytiglossa laevilinguis var. longifolia 257
Justicia riojana Кар oblonga 201
280
Annals of the
Missouri Botanical Garden
Rhytiglossa obtusifolia
Rhytiglossa sei ey var. hirsuticaulis
Rhytiglossa panicu
Rhytiglossa pector alis
Rhytiglossa piahu шеп
Rhytiglossa гере
Rhytiglossa sarmentosa
Ruellia = 'eolata
Sarothec
ак са elegans
Sarothec "ca. glutinosa
Sect. Dianthera subsect. Dianthera
Sect. Dianthera subsect. Saglorith
Sect. Dianthera subsect. Strobiloglossa
267, :
234
Sect. PABA ae
r
Se ricographis
Sericographis macedoana
Sericographis macedoana var. elegans
Sericographis pauciflora
dira didi pauciflora var. speciosor
ericographis squarrosa
Жей
ыша comata
Stethoma pectoralis
Thalestris
Thalestris graminiformis
Jehu yuyo
Vara de la justicia
25:
235, 241
PLANT DIVERSITY OF THE Peter Goldblatt? and John C. Manning?
CAPE REGION OF
SOUTHERN AFRICA!
ABSTRACT
Comprising a land area of ca. 90,000 km”, less than one twe mtieth (5%) the land area of the southern African
subcontinent, the Cape Floristic Region (CFR) is. for its size. one of the world’s richest areas of plant species diversity.
А new synoptic flora for the Region has made possible an accurate reassessment of the flora, which has an estimated
9030 vascular plant species (08.7% endemic). of which 8920 species are flowe n plants (69.5% endemic). The number
of species packed into so small an area is те emarkable for the te mpe rate zor and compares favorably with species
richness for areas of similar size in the wet tropics. The Cape region consists of a mosaic of sandstone and shale
substrata with local areas of limestone. It has a highly dissected, rugged topography, and a diversity of climates with
ү all mostly — in the winter _ and varying from — mm locally to less than 100 mm. Ecological gradients
e steep as a result of abrupt differences in soil. altitude, aspect, and precipitation. These factors combine to form an
ана шек — of local habitats for plants. * —— rived soils have characteristically low nutrient status
and many plants present on such soils have low seed dispersal « spies a jae Pd promoting localized dienten.
, Proteaceae, Restionaceae,
An unusual family ea inc dune Iridaceae. Aizoaceae, Ericac
Rutaceae. and Orchidaceae among the 10 largest familhes i in the flora. following Asteraceae and Fabaceae, as the most
spec iose families. ا radiation has resulted in over spec ies falling in the 10 largest families and
77.4% in the largest 20 families. Twelve genera have more than 100 species and the 20 largest genera contribute some
31% of the total species. Spe cies ric — of the Cape flora is — sized to be the result of ge —— and parapatric
radiation in an area with a mosaic of different habitats due to local soil, climate, and Ken differences that
A
combine to produce steep ecological gradients. Also c contributing to the diversity has been a relatively stable geologic al
history since the end of the Miocene that saw the establishment of a semi-arid and extreme se asonal climate at the
southwestern part of southern Africa
Key words: floristics, Mea type climate, phylogeography. plant diversity, southern Africa, spectation.
Situated at the southwestern tip of the African subcontinent (Goldblatt, 1978, 1997), the Cape
continent between latitudes 31° and 34°30'S (Fig. Floristic Region (CFR) is one of the world’s most
1). the area that has come to be called by biologists — botanically diverse regions. An estimated 9030
the Cape Region has a flora. and to a lesser species of vascular plants (ferns and other vascular
extent a fauna (Stuckenberg, 1962), that is so cryptogams, gymnosperms, and flowering. plants),
sharply distinct from that of the land immediately the majority of which, some 8920 in total, are flow-
adjacent to it that it has impressed naturalists from ering plants, occur there, almost 69% of which are
the time of its discovery by European explorers in endemic (figures based on Goldblatt & Manning,
the 16th century. Indeed, the floristic characteris- 2000, but reflecting taxonomic changes made since
ties of the Cape Region are so unusual that it has the completion of that work). Thus, the flora of the
sometimes been regarded as one of the world’s six Cape Region comprises 44% of the estimated
floral kingdoms (e.g. Good, 1974; Takhtajan. 20,500 species that occur in all of southern Africa
1986). There are. however, no objective criteria for (Arnold € de Wet, 1993; de Wet, pers. comm.). The
distinguishing such “floral kingdoms,” and recog- level of species richness is notable, particularly in
nition of a Cape Floral Kingdom is not universal. Africa, the tropical flora of which is relatively de-
We use the neutral term “floristic region” here sim- paupératé. but is remarkable for the world's tem-
ply for convenience. perate zone, comparing favorably with species rich-
Comprising a land area of ca. 90.000 Кар, less ness for areas of comparable size in the wet tropics
than 5% of the total area of the southern African rather than for areas of temperate climate.
! Fieldwork in southern Africa was supported by the National Geographic Society (grants 5408-95 and 5994-97), the
Missouri Botanical Garden. and the National Botanical Institute. South Africa. We thank Richard Cowling for reviewing
the manuse — and for pro yr numerous helpful comments
2 В. rukoff Curator of African Botany, Missouri НИ Garden, Р.О. Box 299, St. Louis, Missouri 63100,
LS.A. peter.g асро org.
* Compton Herbarium, National Botanical Institute, P. Bag. X7. Claremont 7735, South Africa. manning@nbict.nbi.
ac.Zza
ANN. Missouni Bor. GARD. 89: 281—302. 2002.
Annals
Missouri B Garden
| NS.
T i i
ani
e M
Xe
Figure 1.
a 2000). Abbreviations:
Southwestern; SE, Southeastern.
AP,
Not only is the plant species richness of the
Cape flora exceptional, but its familial and generic
composition is remarkable (Bond & Goldblatt,
1984; Goldblatt, 1997; Goldblatt & Manning,
2000). Unexceptionally for a region of fairly dry
climate, the largest families are the Asteraceae
and Fabaceae, together comprising some 20% of
the total species. The families next in size, how-
ever, are not matched in any other flora—nowhere
else in the world do Iridaceae, Aizoaceae, Erica-
ceae, Proteaceae, and Restionaceae assume such
numerical significance, except in some parts of
Australia where Proteaceae and Restionaceae are
also unusually well represented. Other peculiari-
ties of the Cape flora are the dominance of fine-
leaved sclerophyllous shrubs, the paucity of trees,
and a remarkably large number of geophytes, here
defined as seasonal herbaceous perennials with
bulbs, corms, tubers, or prominent rhizomes (thus
subshrubs that
excluding shrubs and resprout
from a woody caudex, usually after fire). Such geo-
phytes, especially numerous among the monocots
(notably, in order of importance, Iridaceae, Orchi-
daceae, Hyacinthaceae, and Amaryllidaceae), also
include many species of Oxalis (Oxalidaceae) and
Pelargonium (Geraniaceae) as well as other eu-
dicots. Geophytes as so defined comprise slightly
more than 17% of the species in the CFR. Con-
versely, the Cape flora has a surprisingly low pro-
portion of annuals for an area of largely semi-arid
climate. Approximately 6.8% of the species are
annuals, which is a striking contrast to California
Get
aN
HEIGHT ABOVE SEA LEVEL
20 0 20 40 60 80 100km
22°E 23° 24" 25* 26"
The Cape Floristic Region, showing relief, with the phytogeographic centers marked (from Goldblatt &
Agulhas Plain; KM, Karoo Mountain; LB, Langeberg; NW, Northwestern; SW,
(30% annuals)
Kalin Arroyo et al.,
latitude and climate. Both geophytes and annuals
are primarily adapted to seasonally dry climates
Chile (nearly 16% annuals)
1994), areas of comparable
—
and escape the time of year unfavorable for growth
by retreating to underground storage organs or by
ensuring continued survival only by the produc-
tion of seeds.
The compilation of a synoptic flora of the Cape
Region (Goldblatt & Manning, 2000) replaces ear-
lier vegetational analyses based on the work of
Bond and Goldblatt — now very much out of
date. Statistics presented here taken from
Goldblatt and Manning — with minor modi-
fications reflec ‘ting taxonomic changes in press or
published since its completion. Familial and or-
dinal taxonomy is that recommended by the An-
giosperm Phylogeny Group (APG, 1998). Changes
in the CFR made since this synoptic classification
was published are recognition of Veronicaceae
(now including Plantaginaceae and several genera
previously of Scrophulariaceae), enlargement of
Stilbaceae to include non-Cape genera, the reduc-
tion of Achariaceae in Flacourtiaceae, and Prion-
iaceae in Thurniaceae, and the transfer of Hyaen-
anche from Euphorbiaceae to Picrodendraceae
(= Pseudanthaceae) (Olmstead et al., 2001, and
pers. Chase et al., 2000; Savolainen et al.,
00). Tamaricaceae, represented by one species
comm.;
of Tamarix, was omitted in error from the account
of the flora. These changes are represented in the
revised familial statistics.
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
PHYSICAL CHARACTERISTICS
LANDSCAPE AND CLIMATE
Mountain belts of the Cape Region are not par-
ticularly high, generally 1000—2000 m in elevation,
and although the peaks are well below a truly al-
pine zone at the latitude of the Cape region. they
are high enough for winter freezing to be a factor
affecting the vegetation. The mountains are rugged,
and cliffs and exposed rock are evident everywhere.
The rugged topography and vertical landscape am-
plify the effects of local climatic variation with the
result that the mountains offer a greater diversity
of habitats than are present in the lowlands.
The climate is largely mediterranean, and strictly
so in the west. although the eastern half of the CFR
receives substantially more summer precipitation,
Rain thus falls mainly in the winter months and
while summers are hot and dry, they are relatively
less so in the east. In areas of low total rainfall, the
average monthly precipitation distributed in the
summer months may be higher, notably in the Little
Karoo,
evaporation (P/E) ratios effective rainfall is still
but due to more favorable precipitation/
mainly in the winter. South-facing mountain als
benefit from summer moisture in the form of rain
or fog from the southeast trade winds. The narrow
coastal plain in the Knysna area, which has a com-
paratively equable climate and high rainfall. sup-
ports an evergreen, broad-leaved forest. Likewise.
sheltered valleys and locally wet sites throughout
the region, affording higher P/E ratios, support for-
esl vegetation,
Local variation in rainfall is particularly pro-
nounced in mountainous areas, and this is impor-
tant when precipitation is orographic. Mountain
slopes facing prevailing winds receive considerably
more precipitation than those in the lee. Rainfall
patterns in the Cape Region show dramatic varia-
tion in quantity, dropping from 2000 mm per year
on the high mountains of the ranges immediately
facing the coast, to less than 200 mm on the lee-
ward slopes of the interior ranges. Mosaic effects
of soils across the entire region are thus compli-
cated by variation precipitation from the coast
to the interior, as well as changes in seasonality
from the west to the east. In addition, elevation and
aspect affect precipitation depending on the direc-
tion of moisture-bearing winds. The eastern and
western parts of the CFR are considered to be un-
der fundamentally differing climatic controls, prob-
ably a long enduring pattern that has affected the
evolutionary histories of the areas and, hence, re-
sulting in their distinctive floras and faunas (Cowl-
1999).
ing et al.,
The number of ecological niches available to
plant life is multiplied by soil differences, and this
is particularly pronounced as precipitation. levels
drop. With ample rain the effect of soil on vegeta-
Rainfall is
limiting almost throughout the region, however, and
tional composition is less prominent.
vegetation varies conspicuously with soil and avail-
able moisture. Climatic gradients are steep, al-
though perhaps not more so than in most other ar-
eas of mediterranean climate, but the effect may be
compounded in the Cape Region by soil diversity.
Different soil types in the Cape Region support
characteristic vegetation types depending on asso-
ciated levels of precipitation. Forest vegetation is
typical of deeper soils and sites where precipitation
is high and fairly evenly spread. throughout. the
year. As soil qualities change and precipitation be-
comes lower or more seasonal, forest gives way to
shrubby or herbaceous vegetation types. On sandy
soils, forest yields to a sclerophyllous vegetation
(fynbos) in which species diversity decreases and
composition changes until rainfall minimums reach
about 300-250 mm p.a. when a succulent: shrub-
and becomes dominant. On clay soils forest gives
way to fynbos and then to the characteristic renos-
terveld, a shrubland dominated by shrubby, micro-
phyllous Asteraceae. At precipitation levels below
100 mm p.a. renosterveld is increasingly dominated
by succulent perennials. The dissected landscape
ensures that broad sweeps of one vegetation type
are isolated from others by habitats that will not
support their growth. The mosaic of different soil
types alone contributes to increasing diversity, but
the peculiar nature of nutrient-poor soils may result
in more pronounced effects on plant diversity, plant
dispersal, and hence gene flow.
GEOLOGY AND SOILS
Over most of its surface the Cape Region is cov-
ered by soils derived from rocks of pre-Carbonif-
1962), 400 mya
(Fig. 2). Most of these rocks comprise part of the
erous age (King. thus more than
Cape System, an ancient Devonian—Ordovician se-
ries of sediments consisting of alternating lavers of
largely quartzitic sandstones (the Table Mountain
and Witteberg Groups) or fine-grained shales (the
Bokkeveld Group).
surface was folded and warped as Antarctica sep-
During the Jurassic the land
arated from the south coast of southern Africa and
The
folds consistently run parallel to the coasts, result-
South America rifted away from the west coast.
ing in a series of east-west trending mountain rang-
es in the southern half of the Cape Region and
north-south trending ranges in the west.
284
Annals of the
Missouri Botanical Garden
Nieuwoudtville
Ba let. (Sil
Bokkeveld shale (Devonian)
[| Superficial sands and limestones (Quaternary)
; Uitenhage sandstones and shales (Cretaceous)
Witteberg sandstone (Devonian)
Cape
ЖЕ,
d а eae ee PE on
Cape Agulhas
Figure 2.
Differential weathering of the components of the
Cape System has yielded two fundamentally differ-
ent soil types, coarse-grained sandy soils, poor in
essential plant nutrients, and richer, clay soils of
nutrient-intermediate status (Groves et al., 1983).
At the low precipitation levels that are usual in the
Cape Region, these factors become so limiting that
these soils support markedly contrasting vegetation.
Apart from variation in nutrient status, the soils
depart significantly in their structure and water-re-
tention properties. Erosional patterns differ on the
two rock types, and the result is that the mountains
consist primarily of sandstone rocks and the valleys
of shale. Where folding or faulting have been se-
vere, more ancient rocks of the Precambrian Mal-
mesbury Group are exposed. These are largely
shales that give rise to clay soils of the same type
as do the shales of the Cape System. Granitic
schists are locally exposed in deep valleys and
along the west coast, and limestones, mainly of Ter-
liary age, are exposed near the coast where they
are extensive only along the southern coast from
the Agulhas Peninsula east to Mossel B e
coastal plain includes areas with la ш,
soils derived from reworking of Cape Sandstones.
Since these major episodes of folding and rifting in
the Mesozoic, only erosional forces have had a ma-
jor impact on the Cape landscape, modified only
slightly by minor uplift, associated with coastal
downwarping in the mid to late Tertiary (King,
1962). Moving from the coastal plain to the interior,
Geology of the Cape Floristic Region (adapted from Cowling, 1992).
the resultant landscape is a mosaic of coastal lime-
stones and deep sands, or valleys with clay soils
alternating with mountain ranges of nutrient-poor
sands. Local faulting has added a secondary com-
ponent of islands of one rock type embedded in
another. Both the nutrient-poor and nutrient-inter-
mediate soils favor the development of a fairly uni-
form, sclerophyllous, shrubby vegetation that is
fire-adapted (see discussion under Diversity).
WINTER-RAINFALL AND CLIMATIC STABILITY
How important to the flora is the current medi-
terranean climate that prevails over most of the Re-
gion? The vegetation of the Cape Region prior to
the establishment of a winter-rainfall climate was
very different from that now found there. Evergreen
forest has been decreasing since the middle of the
Tertiary, and its diversity has dropped dramatically
since the mid Miocene, some mya, as well
(Coetzee, 1993). Families such as Arecaceae, Cas-
uarinaceae, Chloranthacaeae, Sarcolaenaceae, and
interaceae, no longer found on the African main-
land but still extant in Madagascar, were present in
the Cape Region at least until the mid Miocene
(Coetzee & Praglowski, 1984; Coetzee & Muller,
1985). In addition, early to mid Miocene deposits
on the Cape west coast indicate a fauna adapted to
forest and woodland (Hendey, 1982).
The establishment of the cold Benguella Current
along the west coast of southern Africa in the Mio-
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
cene, with its cooling and drying effects, was prob-
ably the single most important influence affecting
vegetation change in the subcontinent. Summer
drought became increasingly severe in the west as
this current strengthened as a result of the spread
of the Antarctic ice sheet at the end of the Miocene,
ca. 5.5 mya. Even in the later mid Miocene, how-
ever, there was a fairly rich subtropical flora replete
with palms (Coetzee & Rogers, 1982) near Saldan-
ha Bay. on the west coast of the Cape Region in an
area that today supports a largely treeless. succu-
lent or sclerophyllous shrubland. No palm species
occur today in the CFR. Faunal remains suggest
that by the late Miocene the once widespread forest
and woodland were being replaced by more open
savanna (Hendey, 1982). It was probably not until
after the beginning of the Pliocene, i.e.. less than
6 mya, that the present Cape flora т be distin-
guished, although elements of that flora are recog-
nized in Oligocene pollen cores taken within the
Cape Region and nearby (Scholz. 1985). It seems
that
summer drought and lower overall rainfall. was the
clear climatic change, including increased
driving force for vegetational change in southern
Africa into the Pliocene.
Although post-Pliocene changes in climate of the
Cape Region are poorly documented, the climate
appears to have been relatively stable. In compar-
ison with southern Europe. North America, and
southern South America, all of which experienced
cycles of extreme climatic fluctuations with periods
of mild climate alternating with extreme cold and
dryness (Villagrán, 1994), the climate of the Cape
Region appears to have been relatively stable
throughout. the Quaternary. Whereas. in central
Chile. southern Europe, and North America moun-
lain. glaciers developed and winter temperatures
f
their floras could not tolerate, the Cape appears to
=
must have fallen to levels that large portions «
have merely undergone cycles of drier and cooler
The
ameliorating affects of large oceans to the south and
alternating with wetter and warmer conditions.
west would have prevented the extreme conditions
that result in major extinction events. Although no
glaciers developed, there is evidence of colder cli-
mates in the past (Deacon, 1979) consistent with a
temperature depression of the order of 5°C at the
atitude of the Cape Region.
The data of Meadows and Sugden (1991) are
among few documenting the history of the Cape flo-
ra over the past 20,000 years. Their pollen profiles
from the Cedarberg Mountains in the northwestern
part of the CFR show no vegetational changes com-
parable to those known for Chile (Villagrán. 1994)
1978).
or California (Raven & Axelrod, Instead,
there seems to have been a series of subtle shifts
in conditions that favored one community type over
another in the 14,600 years covered in their sam-
pling. The Cedarberg data are especially notable
because that range lies at the northern, more arid
end of the Cape Region, an area that is therefore
particularly sensitive to climatic change. In con-
trast, the somewhat older fossil wood assemblages
from Elands Bay Cave on the west coast of the CFR
(Cowling et al., 1999) document a shift from xeric
to mesic thicket and fynbos and then to afromon-
tane forest at 18,000 BP, indicating higher moisture
conditions in the west during the Last Glacia
In the southern Cape the lowering of sea €—
as much as 120 m below present levels at times
during the Pleistocene resulted in the extension of
the coastal plain off the southern coast of Africa.
The vegetation along the coast at this time was
probably grass-dominated, and it supported the
dominant alcephaline and equid fauna (Klein,
1977).
this area has a large grass component.
Even today, the vegetation on clay soils of
FLORISTIC COMPOSITION
MAJOR FAMILIES AND GENERA
The Asteraceae, usually the largest family in flo-
ras of arid to semi-arid regions, are also the most
speciose family in the Cape flora, with 1030 species
Table
since the flora was last documented (Bond & Gold-
blatt, 1984) now make this family the second larg-
Additions to several genera of Fabaceae
—
est in the flora (previously believed to be fourth
argest), which is also unexceptional, as Fabaceae
are well developed in most parts of the world. How-
ever, the huge contribution made by Iridaceae, Ai-
zoaceae, and Ericaceae, next in numbers of species
(Table 1), is a unique aspect of the flora (and con-
sequently of the southern African flora as a whole).
Scrophulariaceae, Proteaceae, and Restionaceae
follow in importance, showing a pattern, described
in more detail by Goldblatt and Manning (2000).
Aizoaceae, with its huge southern African. repre-
sentation of Mesembryanthemoideae, are probably
the second largest family in the southern African
flora (Goldblatt, 1978) and appear to be the largest
family in the Namaqualand—southwestern Namibia
region (R. Cowling, comm.) that lies to the
north of the CFR and also has a winter-rainfall and
summer-dry climate. This arid zone has such strong
floral affinities to the Cape flora that the inclusion
of Namaqualand-southwest Namibia in the CFR to
pers.
comprise a Greater Cape flora has been given se-
rious consideration (Bayer, 1984: 1991,
1997). The relationships of the extended Namaqua-
Jürgens,
286
Annals of the
Missouri Botanical Garden
Table |
Ranking of the 20 largest families in the Cape Flora Region as indicated by species number. Family
circumscriptions reflect the recommendations of the Angiosperm Phylogeny Group (1998). These families contribute
6989 species to the flora, or 77.4% of the total.
Number endemic
(% of total
of tot Total g a
Family Total species species) (number endemic) Species/genus
l Asteraceae 1036 655 (63.2) 121 (33) 8.6
2 Fabaceae 761 629 (82.7) 37 (6) 20.6
3 Iridaceae 677 540 (79.8) 28 (6) 24.2
1 Aizoaceae 659 524 (79.5) 76 (18) 1.5
5. Ericaceae 657 637 (96.9) 1 (0) 651
6. Scrophulariaceae 414 297 (71.7) 33 (7) 12.5
da Proteaceae 329 319 (97.0) 14 (9) 23.5
8. Restionaceae 318 294 (92.5) 19 (10) 16.7
9. Rutaceae 273 257 (94.1) 15 (6) 18.2
10. Orchidaceae 227 138 (60.8) 25 (2) 9.1
11. Poaceae 207 80 (38.6) 61 (3) 3.4
12: Cyperaceae 206 101 (49.0) 29 (4) 7.1
13. Hyacinthaceae 191 83 (43.5) 14 (0) 13.6
14. Campanulaceae 183 140 (76.5) 13 (6) 14.1
15. Asphodelaceae 157 81 (51.6) 8 (0) 19.6
16. Geraniaceae 157 9] (58.0) 3 (0) 52.3
17. Polvgalaceae 141 122 (86.5) 3 (0) 47.0
18. Rhamnaceae 137 127 (92.7) 5 (1) 27.4
19. Crassulaceae 134 35 (26.1) 5 (0) 26.8
20. Thymelaeaceae 125 94 (75.2) 4 (1) 31.3
Y = 6989 5244 (77.0) 514 (112)
land flora (or Succulent Karoo Region) seem un-
questionably closer to the Cape flora in its tradi-
tional sense than to the flora of the summer-rainfall
karoo (Nama-Karoo Region) although their common
boundary is not clearly fixed and there is evidence
for its east-west shift in the past (Jürgens, 1991).
The large numbers of species of Proteaceae and
Restionaceae, seventh and eighth largest families,
respectively, are another striking feature of the flo-
ra. The importance of Ericaceae, Proteaceae, and
Restionaceae both in terms of biomass and species
diversity is widely appreciated, but the huge num-
ber of species of Iridaceae, predominantly a family
of herbaceous, seasonal geophytes, is especially no-
table. Nowhere else in the world does this family
comprise such a significant floristic component. In-
deed, the adaptive radiation of Ericaceae and Iri-
daceae in the Cape flora is one of its most striking
features. The diversification of Ericaceae, Protea-
ceae, Restionaceae, and even Cyperaceae, is close-
ly associated with the impoverished sandstone soils
of the Cape mountain ranges, and these families
are poorly represented on other soils. The massive
radiation of Fabaceae and Iridaceae shows no such
correlation. Although the wealth of Scrophulari-
aceae, sixth largest family in the Cape flora, seems
remarkable in a world context, the family is also
well represented across Africa, especially in the flo-
ras of drier areas (Maggs et al., 1998). In the Cape
flora Scrophulariaceae contribute 166 species to
the annual flora, far more than does the next most
important family, Asteraceae, with 138 annual spe-
cies. As circumscribed for the Cape flora (Goldblatt
& Manning, 2000) Scrophulariaceae do not include
the parasitic and e genera now referred
to Orobanchaceae (A 98) but, nevertheless,
it remains a major family in the flora. Rutaceae,
ninth largest family, are also a surprising aspect of
the Cape's floristic composition. Over 95% of the
273 species of Rutaceae there are small shrublets
belonging to the tribe Diosmeae, most members of
which are restricted to the Cape flora, and reflect
the large numbers of shrub species in the flora. The
wealth of Polygalaceae, Rhamnaceae, and Thyme-
laeaceae, 17th, 18th, and 20th in size, respectively
(Table 1), likewise exemplify the importance of the
shrubby habit in the Cape flora
Poaceae are comparatively poorly represented in
the Cape flora. Although they are the third largest
family in the flora in number of genera (Table 1),
they are only llth in size in total species, with
fewer representatives than Restionaceae and barely
more than Cyperaceae, the two other families gen-
erally that occupy similar habitats. This situation is
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
families of the CFR. Note that the
characteristic and nearly endemic
Table 2. Endemic
family Bruniaceae has
11 genera and 61 endemic species of a total 64 in the
family and is not included here, its range extending east-
ward to southe rn KwaZulu-Natal. In addition, Olmstead et
(2(
=
=
~
—
) have proposed enlarging Stilbaceae to include
non Cape genera rendering it no longer endemic.
Number
of genera
Number
Family of species
Penaeaceae 7i 23
Sulbaceae s. str. 5 14
Grubbiaceae 1 8
Roridulaceae l 2
Geissol l l l
Totals 15 13
paralleled only in southwestern Australia but con-
trasts sharply with adjacent southern Africa where
Poaceae are prominent and diverse. Poaceae are
the largest family in the flora of Namibia (Maggs et
al., 1998) and one of the five largest in southern
Africa excluding the CFR (Goldblatt, 1978).
Although 150 families of seed plants and another
23 of ferns and other vascular eryptogams are rep-
resented in the flora, remarkably few account for
the bulk of the species. While 23 families in the
flora have over 100 species, only 12 have over 200
species. In contrast, 38 families have just one spe-
cies each. The largest 10 families account for 5351
species, well over half the total 9030 in the flora,
and the largest 20 families account for 6989 spe-
cies, over 77% of the flora (Table 1
ENDEMIC FAMILIES
The unique floristic composition of the Cape Re-
gion, with its high representation of Ericaceae, Ir-
idaceae, Proteaceae, and Restionaceae (Table 1). is
emphasized by the presence of several families that
are endemic or nearly зо. The endemic families are
all eudicotyledons of diverse affinity and relatively
low evolutionary specialization (Table 2). The larg-
est is Penaeaceae (Myrtales), followed by Grubbi-
and
aceae (Cornales). Roridulaceae (Ericales).
Geissolomataceae (together with Ixerbaceae and
Strasburgeriaceae, sister to Crossosomatales) (clas-
sification following APG, 1998). Based on a molec-
ular clock calibrated using 135 mya for the diver-
gence of the eudicot lineage (V. Savolainen et al.,
2000.
have diverged 20 mya from its sister clade, the Af-
and unpublished data), Penaeaceae may
rican Oliniaceae plus the Neotropical Alzateaceae,
while Roridulaceae diverged from Ericaceae, its
il., 2000), са. 48
closest relative (Savolainen et
a
mya. Geissolomataceae appear older, having di-
verged perhaps 55 mya from Ixerbaceae plus Stras-
burgeraceae. Grubbiaceae may have diverged from
the
Tertiary, 63 mya. We have no comparable data for
Cornaceae plus Hydrostachyaceae in earliest
the monotypic, near endemic monocot family Lan-
ariaceae. Stilbaceae (including Retziaceae) (Lami-
ales) are provisionally regarded as a Cape endemic
eg)
amily, but if the changes to its circumscription sug-
gested by Olmstead et al. (2001, and pers. comm.)
are accepte d, the family occurs across sub-Saharan
Africa,
(see below).
Arabia, the Mascarenes, and Madagascar
Bruniaceae, one of the distinctive families of the
Cape flora, are almost endemic. Of an estimated 64
species in ll genera, just 3 species in 2 genera
extend outside the confines of the Cape Region, 2
locally, and 1 as far east as southern KwaZulu-Na-
tal. Bruniaceae are now thought to be the sister
group to the order (Savolainen et al..
2000).
rank.
Dipsacales
perhaps meriting recognition at ordinal
The discovery of pollen matching modern
Bruniaceae in early Tertiary and late Cretaceous
(?Senonian) deposits in northern Namaqualand (S.
E. de Villiers,
the CFR, attest to considerable age for the
pers. comm.), well to the north of
—
amily
in southern Africa. The pollen record also accords
with Savolainen’s preliminary early ‘Tertiary dating
of the divergence between Bruniaceae and Dipsa-
cales at about 57 mya (V. Savolainen, unpublished
data).
Retziaceae (1 genus: | species) have often been
onsidered an endemic Cape family (e.g.. Bond &
Goldblatt,
been in dispute (Goldblatt et al..
1984), although its affinities have long
1979). DNA se-
quence analysis shows the genus nested in Stilba-
1994; Savolainen et al., 2000).
Floral specialization for bird pollination appears to
The
familial status of Stilbaceae is not in question, but
Olmstead et al. (2001) included the Afro- Arabian
Vuxia in tribe Stilbeae and added the Afro-Mada-
Halleria the African
Bowkerieae to the family (previously Scrophulari-
ceae (Bremer et al.,
be the source of most of its distinctive features.
gascan and southern tribe
aceae) rendering Stilbaceae no longer endemic to
the CFR or even to southern Afric
In contrast to the Cape Region, which alone has
5 (or more likely 4) endemic families, all of south-
ern Africa perhaps has just 10 endemic families (9
according to the revised concept of Stilbaceae). In
addition to those absolutely restricted to the Cape
Region, the southern African endemic families are
the eudicots Bruniaceae (ordinal position sister to
Dipsacales, 11: ca. 64), Greyiaceae (Geraniales,
З), and Rhynchocalycaceae (Myrtales, 1:1); the
288
Annals of the
М Botanical Garden
Table 3.
9030 species occur in the flora, of which 6208 are endemic
Ranking by numerical size of the 20 largest ger
vera in the CFR (endemic species number). An estimated
(68.7%
Erica 657 (637)
spalathus 272 (258)
Pelargonium
thosme
Total in largest 10 genera = 1935 spp. (21.4% of flora)
Combined total in largest 20 genera = 2851
spp. (31.6% of
Muraltia 106 (100)
Gladiolus 106 (86
Selago 100 (77
Crassula 95 (26
Disa 92 (78
Ruschia 88 (79
estlo 85 (82
Leucadendron 82 (79
Helichrysum 81 (34
Thesium 81 (35)
flora)
monotypic monocot genus Lanaria is also currently
regarded as comprising its own family, Lanariaceae
(Asparagales) (APG, 1998; Chase et al., 2000); and
the cycad family Stangeriaceae (1:1). Lanaria is
widespread in the Cape Region and extends outside
Aitoni-
Curtisiaceae, and Oftiaceae have at times
its confines a short distance to the east.
aceae,
been accorded recognition but they are no longer
considered to be separate families. They are readily
referable to Meliaceae (Pennington & Styles, 1975),
1993), and Scrophulari-
1979), respectively. Prioniaceae
Cornaceae (Xiang et al.,
aceae (Goldblatt,
(Poales), treated as a family by Munro and Linder
(1998) for the monotypic and largely Cape Prion-
ium, and recognized by Goldblatt and Manning
(2000) for the Cape flora, are now regarded as be-
longing to Thurniaceae (Chase et al., 2000). Behn-
iaceae, described for the monotypic southern Afri-
can Behnia, is an endemic southern African family
(1997), although it is
reported to occur in Zimbabwe. However, the status
according to Conran et al.
of Behniaceae is in question and it is likely to be
subsumed in a more widely circumscribed Agava-
ceae (M. W. Chase, pers. comm.). Likewise, Achar-
iaceae (Malpighiales, current name for Flacour-
2000), two species
of which also occur in the Cape Region, are often
tiales, 3:3) (Savolainen et al.,
regarded as a southern African endemic family. The
genera are, however, nested in Kiggelariaceae (Sa-
volainen et al., 2000) and are regarded here as
members of that family.
GENERA
Some 944 genera of seed plants (or 990 genera
of vascular plants), about half of those occurring in
southern Africa, are represented in the Cape flora,
of which some 160 or 16.3% are endemic (Gold-
blatt & Manning, 2000). The level of generic en-
demism is modest and reflects little of the unusual
nature of the flora. The number of near-endemic
genera (those of moderate size with just one or two
species extending locally outside the Cape Region)
is, however, dramatically high. The largest genus in
the flora by far is Erica, with some 657 species
lable 3). Changes in the circumscription of Erica
—.
Oliver, 2000) have now resulted in the inclusion
—
of all the minor genera of southern African Erica-
ceae: Ericoideae, leaving Erica with over 7% of the
species in the entire flora. It is unclear whether this
remarkable pattern of speciation without generic
diversification is associated with the relatively re-
cent arrival of ancestral ericaceous stock in the
CFR, or with adaptive radiation following the es-
tablishment of a semi-arid climate there. By com-
parison, the smaller families Proteaceae and Res-
tionaceae appear to belong to old African (or even
Gondwanan) groups, now poorly represented else-
where in Africa, and they show the highest levels
of endemism at the generic level. These two fami-
lies plus Bruniaceae are the only non-endemic fam-
ilies that show greater than 50% generic endemism.
Aspalathus (Fabaceae) is the second largest ge-
nus, with followed by gees
Phylicc
Rhamnaceae), Lampranthus (Aizoaceae), and *
2 species,
—
Geraniaceae), Agathosma (Rutaceae),
—
alis (Oxalidaceae), each with between 118 and 148
species (Table 3). Thirty-six genera have over 50
species and 13 genera have over 100 species. The
1935
The next 10 largest genera
10 largest genera contribute over 21%, or
species, to the flora.
contribute an additional 922 species. The 20 larg-
est genera in the Cape flora thus contain over 31%
of the total species.
There is no one unifying ecological pattern evi-
dent in the species-rich genera. Agathosma, Aspa-
lathus, Cliffortia, Erica, Phylica, and the two larg-
est genera of Proteaceae, Leucadendron and Protea,
are best developed on sandy soils and are most
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
Table
The d represent the percentage of the total flora. The
Selected statistics for the CFR and various comparable regions (emended from Bond & Goldblatt, 1984).
ten largest genera were determined by species number.
Mean sp.
number
10 largest
Percentage of total flora
Annual
Region per genus genera Monocots Asleraceae species
Cape flora 9.] 21.4 24.5 11.5 0.7
Southern Africa (incl. Cape flora) 9.6 15.1 23.0 11.0 7.0
Cape Peninsula 1.2 17.5 34.6 11.5 9.6
Nata 3.9 17.0 27.1 11.4 ca. 0.5
Eastern North America 5.2 21.8 28.2 12.7 8.7
Europe 1.0 14.0 183.0 12.0 ?
California Flora 5.2 15.2 19.2 13.6 27.4
Sonoran Desert 3.3 12.8 12.1 15.0 21.4
Texas 3.9 10.2 24.4 13.4 20.4.
Hawaii 1.4 81.0 11.8 15.9 0.04
New Zealand 5.1 26.3 27.9 12.5 6.0
diverse in montane habitats. In contrast, species of ing that elsewhere in the world, is thus no more
Lampranthus, Moraea, Pelargonium, Oxalis, Glad-
iolus, and Crassula appear to occur with equal fre-
quency on nutrient-poor, nutrient-intermediate, or
comparatively rich soils and favor lowland habitats.
Lampranthus and Crassula are succulents, and
Disa, Oxalis, Gladiolus,
geophytes, as are some species of Pelargonium. The
and Moraea are seasonal
remaining genera among the largest 20 comprise
mostly shrubs or small trees. A few species of Pel-
Crassula, and Helichrysum are
argonium, Senecio,
annual species. Thesium species are hemiparasitic
shrubs.
The most obvious shared factors in diversified
genera in the Cape flora seem to be either a shrub-
by habit or seasonal geophytism. Over 17% of the
total species in the flora are geophytes with bulbs.
corms, tubers, or rhizomes. The number of species
with underground perennating buds would be even
higher if plants with woody caudexes were regarded
as geophytes.
The ratio of species per genus, 9.1 (9.3% ex-
cluding ferns) in the Cape flora (Table 4). is partic-
—
ularly high compared to other floras and is one o
the highest in the world (Fenner et al., 1997). al-
though southern Africa as a whole. including the
Cape Region, has a comparable ratio (9.6%, fide
Goldblatt, 1997). A ratio of three dicots to one
monocot species in the Flora is close to the average
for floras across the world. Although the proportion
of monocots does not seem unusual, the monocot
families that are represented in the Cape flora are
most. distinctive. Some half of the species in the
monocot families are geophytic and belong in the
petaloid monocot families, notably Iridaceae, Or-
and Amaryllidaceae.
chidaceae, Hyacinthaceae,
The proportion of monocot to dicot species, match-
than coincidence.
PALEOENDEMIC GENERA
The endemic and near-endemic families of the
Cape Region are all small and contribute relatively
few genera and species to the flora, and they are
best regarded as paleoendemics. With the excep-
tion of the rhizomatous perennial Lanaria, members
of these families are all evergreen, sclerophyllous
shrubs. They are often summer-flowering. usually
have small flowers, and typically grow on sand-
stone-derived soils. We speculate that these plants
are relicts of an ancient temperate southern African
flora adapted to nutrient-poor soils.
Among non-endemic families, these paleoen-
demic genera are often members of the geologically
oldest communities, tropical thicket and evergreen
forest. Especially notable are Platylophus, a mono-
typic genus of Cunoniaceae and one of two conti-
nental African members of this family. The other,
Cunonia capensis, is common in the Cape Region
Africa. Other
monotypic genera such as Heeria and Lauroph yllus
but extends into eastern. southern
(Anacardiaceae), Hartogiella and Maurocenia (Ce-
lastraceae). Lachnostylis (Euphorbiaceae),
Smelophyllum (Sapindaceae) also exemplify the pa-
leoendemic component of the depauperate tree el-
and
ement of the flora and mostly have ranges restricted
to the southern or eastern portion of the Cape Re-
gion (Goldblatt & Manning, 2000)
In non-forest habitats there are few paleoendem-
ic genera or species apart from members of the en-
demic (and near-endemic) families of the Cape Re-
include the monotypic shrublets.
gion. These
Empleuridium (Celastraceae) and Lxianthes (Scro-
Annals of the
Missouri Botanical Garden
Table 5.
flora, central Chile, and the CFR. Figures are percentages
Comparison of life forms in the California
of total species number; perennials below include geo-
phytes and ее нта Data for California and Chile are
from Kalin Arroyo et al. (1994).
Percentage of total flora
Region Trees Shrubs Perennials Annuals
Cape Region 2.5 53.3 37.5 6.7
California 4.6 11.0 56.2 30.2
Central Chile 2.9 17.8 63.4 15.8
phulariaceae or Stilbaceae, Olmstead et al., 2001).
The small tree, Hyaenanche (Euphorbiaceae or Pi-
crodendraceae, Savolainen et al., 2000), also mono-
typic, and Metrosideros angustifolia (Myrtaceae), a
member of an otherwise Australasian genus, show
an odd pattern for the tree flora. Their ranges are
restricted to the western half of the Cape Region
where there are few tree species. Metrosideros an-
gustifolia, the only African member of Myrtoideae:
Metrosiderinae, seems as geographically isolated
from its closest relatives as Cunonia and Platylo-
phus. Like Metrosideros, Bulbinella is also a Cape—
Australasian disjunct, but in this case the radiation
within the genus has occurred largely in the Cape
Region. The dwarf, tree-like monocot Prionium
(Thurniaceae), «
sandstone soils, is widespread in the Cape region
The other
f watercourses in nutrient-poor
~
but extends some distance to the east.
member of the family is the Brazilian shield genus
Thurnia, the pair thus exhibiting an unusual dis-
junction. The taxonomically isolated Oldenburgia
(Asteraceae-Mutisieae), with three Cape species
and one occurring a short distance beyond its east-
ern limits, is perhaps another example of that dis-
tribution pattern, for its closest allies occur in the
Guyana Highlands. The small number of paleoen-
demic taxa emphasizes the huge contribution that
recent speciation in a narrow range of families and
genera has made to the total species diversity in
the Cape Region.
LIFE FORMS
In contrast to other mediterranean floras, the
Cape flora has relatively few trees, and this life
form only accounts for some 220 species, about
2.596 of the flora. California flora ac-
count for 4.696 of the species, but in Chile some
2.996 of the species are trees (Table 5). Most of the
Cape flora are shrubs
Trees in the
remaining elements of the
and perennial herbs. The shrubby habit is the most
common life form in the Cape Region, accounting
for an estimated 4797 species, 53.3% of the total
flora. Shrubs are diverse in form, but typically in-
clude species with sclerophyllous, and mostly mi-
crophyllous leaves, the characteristic that has given
rise to the word fynbos, an Afrikaans word describ-
ing fine-leaved vegetation. Shrubs also include
large numbers of species with succulent leaves (es-
pecially Aizoaceae). Stem-succulents include spe-
cies of Apocynaceae and Euphorbiaceae, some of
which are so reduced in size that the term shrublet
hardly seems appropriate. The Cape flora stands
out when compared to that of both California and
Chile in the overwhelming proportion of shrubs (Ta
ble 5). which is largely explained by the nutrient-
poor soils that favor this life form.
The Cape flora shows some notable differences
with other mediterranean floras. One of these is a
surprisingly low proportion of annuals—only some
609 species, about 6.796 of the total flora, as com-
pared to 30% and 15% respectively for California
and central Chile (Table 5) (Kalin Arroyo et al.,
l 1990). No comparable figures
are available for Basin. Al-
though the proportion of annual taxa in the CFR is
ow, the annual flora is quite species rich. The total
»
Nx
4; Cowling et al.,
the Mediterranean
number of annual species is actually almost twice
as high as the 378 annuals recorded in central
Chile, an area of comparable size, and although the
Cape has less than half the 1279 annuals in Cali-
fornia, its area is only about one fourth that of Cal-
ifornia. For its geographic area then, the Cape flora
is not depauperate in annuals, but its wealth of
other life forms makes the annual habit appear un-
der-represented. A small annual flora is also char-
acteristic of southwestern Australia, an area that
has a recent geological history comparable to that
of the Cape Region and a similar pattern of nutri-
ent-poor sandstone soils and richer clay. The low
proportion of annuals in the Cape flora has re-
mained without a satisfactory explanation since it
was first noted by Bond and Goldblatt (1984) but
the answer may simply lie in the disproportionate
numbers of other life forms, especially microphyl-
lous shrubs that are particularly well adapted to the
nutrient-poor soils.
Two families, Scrophulariaceae and Asteraceae,
are especially important in their contribution to the
annual flora. Scrophulariaceae, with 166 species,
contribute the largest number of annuals, and not,
as might be expected, Asteraceae, which have some
138 annual species (Table 6). The Aizoaceae, Bras-
sicaceae, Campanulaceae, Crassulaceae, Cypera-
ceae, Fabaceae, and Poaceae, each contribute be-
20 and to the annual flora.
mlaceae, in particular,
tween 35 species
need taxonomic
pal
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
291
Table 6. Families with more than 10 annual species
in the CFR. Total: 609 species (6.7% of the flora).
Table 7. Total species (percentage in parentheses) of
the different life forms in the Cape flora.
Family Species number
Scrophulariaceae 166
Asteraceae 138
Campanulaceae 35
Brassicaceae 33
Poaceae 31
\izoaceae 25
Fabaceae 25
Crassulaceae 23
Cyperaceae 21
Gentianaceae 18
Molluginaceae 16
study and our estimation of the number of species
in the family, including its annual component, 15
subject to significant revision.
In contrast to the low proportion of annuals, the
Cape Region has perhaps the highest proportion of
geophytes of any part of the world, and is four to
five times richer in this life form than is docu-
mented for other mediterranean floras (Esler, 1998:
Esler et al., 1998; Goldblatt € Manning, 2000). At
least 1550 species, over 17% of the total, have spe-
cialized underground organs. including bulbs.
corms, rhizomes, or tubers and are seasonally dor-
mant (Table 7). The overwhelming number of geo-
phytes are monocots, with over 1300 geophytic spe-
cies, 662 of which belong to one family, Iridaceae.
Most of these geophytes are seasonal and lie dor-
mant underground in the dry season, but we have
Species Species % of
Life form number total flora
*Perennials 1025 (11.4)
‘Trees 221 (2.4)
Shrubs 1805 (53.2)
Geophytes 1575 (17.4)
Graminoids 795 (8.8)
Total, all perennials 8421 (93.3)
Annuals 609 (0.7)
Total 9030 (100)
* Perennials excluding trees. shrubs, geophvtes. and
graminoids.
included in the geophyte category the few more or
less evergreen species (e.g.. Agapanthus, Kniphofia)
with similar underground organs. The other main
category of the monocots are the graminoids. that
is, the perennial species of Cyperaceae, Juncaceae.
Poaceae, Restionaceae, and a few other families.
which account for 795 species (8.8% of the flora).
VEGETATION
Far from having uniform vegetation, the Cape
Region encompasses five biomes (Fig. 3) and sev-
eral distinctive vegetation types. each with its own
suites of species and physical characteristics
(Rutherford & Westfall, 1994; Cowling & Holmes,
1992a). The most common and distinctive biome is
=,
heathland, locally called fynbos, an analogue «
E (0 Fynbos Karroid shrubland
X ES Renoster Shrubland E Dune thicket
м Afromontane forest
Figure 3.
Major vegetation communities of the Cape Floristic Region (adapted from Louw & Rebelo, 1996).
292
Annals of the
Missouri Botanical Garden
Table 8.
Comparison of species richness (indicated by — number per area), endemism, and the proportion of
life forms in the floras of the six phytogeographic centers of the
e flora (n/a = not available). For comparison, data
are listed for the Cape Peninsula, a small area within the — Center.
Floristic center Physical Total % inm Geophyte
(listed from west area species Species/ species Tree species
to east 10% km? number 10% km? endemic species Nein (90) number (9€)
Northwest Center 22 4066 184.8 26.3% 69 415 (10.3) 55 (21.2)
Southwest Center 23 4661 202.7 32.0% 95 312 (6.8) 846 (18.2)
Agulhas Plain 3 1374 158.0 14.9% 24 92 (6.7) 202 (14.7)
Karoo Mountain 19 2151 113.2 15.5% 47 130 (6.1) 330 (15.5)
Langeberg Center 7 2363 337.6 11.7% 100 127 (5.4) 389 (16.4)
Southeast Center 18 2832 157.3 9.7% 163 156 (5.5) 427 (15.1)
Cape Peninsula 4.7 2250 178.7 1.59€ n/a n/a n/a
Cape Region 90 9030 100.3 08.7% 221 609 (6.7) 5 (17.4)
Californian chaparral and Mediterranean maquis.
Shrubs with ericoid or short, narrow, often needle-
like leaves predominate, but most species of Pro-
teaceae, a family common in this vegetation, have
broad sclerophyllous leaves. Fynbos typically oc-
curs on sandstone soils. A second distinctive veg-
etation type, renosterveld, is usually restricted to
richer fine-grained soils. It shares few species with
fynbos although they often grow adjacent to one
another. Microphyllous Asteraceae are common in
renosterveld, which consists of a dense shrubland
with a rich herbaceous understory that becomes ev-
ident after fire or clearing but is often suppressed
under a mature shrub cover. Dry sites with rainfall
less than 200 mm p.a. support a vegetation of small
succulent-leafed shrubs, including many Aizoaceae
called succulent karoo.
Thicket (a dense, semi-succulent and spinescent
and Asteraceae, in a biome
evergreen shrubland to low forest) and evergreen
forest make up the remaining biomes.
Fire is an integral part of the ecology of the Cape
Region and accounts for several aspects of the veg-
etation (Cowling, 1987). Growth form in mature fyn-
bos and renosterveld is a relatively uniform, closed,
low canopy of twiggy and microphyllous to sclero-
phyllous shrubs. These vegetation types are highly
prone to periodic fire. Fire itself has a disruptive
effect on the vegetation. It has obviously been a
feature of the ecology for so long that there is a
large flora of ephemerals, geophytes, other peren-
nials, and short-lived shrubs that appear in the
years following a fire, often flower profusely, and
subsequently disappear, as they are succeeded by
longer-lived shrubs. The long-term ecological con-
sequence of fire on the flora is the existence of a
niche for species that grow rapidly after fire to per-
sist and bloom in the immediate post-fire years.
This fire-adapted suite of species contributes sub-
stantially to the overall diversity in the flora. Ma-
ture vegetation is affected by fire in more subtle
ways, but fire may cause local perturbations in spe-
cies Composition and the elimination of some taxa.
DIVERSITY
REGIONAL DIVERSITY
The patterns of endemism within the Cape Re-
gion are fairly consistent, and examination of these
patterns in selected genera that have diversified
largely on sandstone substrates has resulted in the
recognition of several regional centers of endemism
(Fig. 1, Table 8). Weimarck (1941) pioneered this
field, which has now been refined by Cowling and
his coworkers (e.g.. Cowling & Holmes, 1992a
Cowling & McDonald, 1999). The presence of these
centers suggests that exchange between them is
limited because of effective geographic isolation or
because different microclimates in each center fa-
vor local species at the expense of migrants.
One aspect of our account of the Cape flora
(Goldblatt & Manning, 2000) has been the formal
recognition of phytogeographic centers so that dis-
tributional data for species can be analyzed within
е ius, we have been able to estimate for
the first time the floristic diversity for each center.
To some extent the statistics are approximate be-
cause some centers are under-collected (or under-
cited in taxonomic accounts). We suspect that the
Karoo Mountain (KM) and the Agulhas
Plain Center (AP) have more species (and thus low-
er levels of endemism) than our data suggest
Center
—
Fig.
E
—
At the geographic center of the Cape Region, the
Southwestern Center (SW) has the largest number
of species (4661) and the highest level of endemism
(32%). The Northwestern Center (NW) follows in
taxonomic size (4066 species) and endemism
(26.3%).
The Karoo Mountain and Southeastern
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
Centers have much smaller floras and substantially
lower levels of endemism. Much smaller in extent,
the Langeberg (LB) and Agulhas Plain (AP) Centers
understandably have smaller floras. The Southeast-
ern Center (SE), of almost the same physical size
as the NW and SW Centers, has a markedly smaller
flora (2832 species),
This may be explained by its more equable and
and only 9.7% endemism.
apparently less diverse climate or simply by higher
levels of extinction during colder and drier periods
Differences in levels of ende-
The SW and
NW Centers each have about twice the proportion
of the Pleistocene.
mism across the Centers are striking.
of endemic species as the others, which may be a
reflection of their greater climatic diversity.
The different life forms are unevenly distributed
across the Cape Region (see Table 8), and the num-
bers of species of the two most distinctive Ше
forms, annual species and geophytes, drop dramat-
ically moving from west to east. The summer-dry
NW and SW Centers have the largest numbers of
geophytes and the highest proportion of geophyte
species in the Cape flora, 21.2% and 18.2%. re-
spectively. These two centers each have over 50%
of all the geophytes in the entire CFR. In compar-
the other centers have between 14.7% and
16.4% geophytes. Annual species are more. com-
mon in the west, and the NW Center has 10.3%
annual species. while the remaining Centers each
The NW Center
alone has 65% of the total annual species in the
ison,
have no more than 6.8% annuals.
Cape flora. The distribution of trees shows the con-
verse, with relatively few tree species in the NW
Center and the highest numbers by far in the SE
Center, which may be explained by its less pro-
nounced seasonality and more predictable rainfall.
Cowling (pers. Comm.) suggests that proximity to
the source pool of species for recolonization during
warmer (e.g.. Holocene) periods is probably the
most important factor in explaining tree distribu-
tions.
These patterns seem directly related to climate.
Both geophytes and annual species seem best
adapted to a seasonally extreme climate with a wet
winter and dry summer. A climate with higher, sea-
sonally more evenly distributed rainfall, character-
of the LB and SE
favor a tree flora and fewer geophytes and annuals.
Istic Centers, seems likely to
Like annuals. geophytes are adapted for survival in
semi-arid, seasonal habitats. This is reflected in the
grealer representation of annuals and geophytes in
the western half of the Cape Region, where summer
precipitation is lowest. The unexpectedly high
numbers of trees in the SW Center reflect the dis-
sected landscape with the presence of fire-sheltered
valleys and the regular occurrence of rainfall in the
summer along its southern coast and interior.
COMPARISONS WITH OTHER FLORAS
An aspect of the Cape flora that is of particular
interest is the high level of species diversity, both
For its size (ca. 90.000 kn).
the number of species of vascular plants in the
CFR. 9030 (8918 seed plants plus 112 pterido-
phytes), is comparable with areas of the wet Neo-
tropics (Table 9). Thus Panama (75.000 km?) has
7300 seed plant species and Costa Rica (54,000
Table 9). In
fact, southern Africa as a whole has a particularly
regional and local.
km?) may have over 9000 species (see
rich and diverse flora for a predominantly temper-
ate region. The area customarily treated for floristic
purposes as southern Africa (Botswana, Lesotho,
Namibia, South Africa, and Swaziland) has about
20.500 native vascular plant species in an area of
2.074.000 km?, and South Africa alone may have
18.275 seed plants plus
1.221.000 km? (€. de Wet.
This is striking compared. with an
estimated 19,000 species in all of North America
north of Mexico (19,341,000 km). or the estimated
800 seed
plants plus ca. 700 pteridophytes) recognized for
Peru. an area of 1.285.000 km? (Brako & Zarucchi.
1993:
gional context, all of tropical Africa may have about
26.500 species (Lebrun & Stork, 1997), in an area
nearly 10 times larger than that of southern Africa
some 18,500 species (ca.
245 pteridophytes) in
pers. comm.).
16.500 native vascular plant species (15.
Zarucchi, pers. comm.). To put this in a re-
and about 250 times as large as the Cape Region.
Africa and North Africa have
mately 21,500 additional species (those not shared
Southern approxi-
with adjacent tropical Africa). making a total of ca.
47.000 species for the entire African continent. The
tiny Cape Region, less than 0.5% of the total area
of Africa, then has about one fifth of all the species
on the continent. Subtropical southern Africa, ex-
cluding the Cape Region, has only about 14.300
species, a figure comparable with that for Tropical
For the Af-
rican continent then, not only is the Cape Region
—
T
last Africa (Polhill, unpublished data).
remarkably rich in species, but southern Africa has
a higher species diversity than would normally be
predicted given the general trend that species di-
versily increases toward the equator.
The species richness in the Cape flora is. by any
measure, remarkable. Moreover, some 0208 or
about 68.7% of the species are endemic there (Та-
ble 9). The high degree of species endemism in the
Floristic
Cape flora compared to the California
Province. for example, with some 4240 species.
Annals of the
294
issouri Botanical Garden
M
9'£0I cel Qr вә OOLZ e2 -— = 97 ALIS
єз él 9 “Bo TS0c — — VE втшрлес
и 48! él Gol 2 00? — — Ic asauuodo[aq
90681 él OT F2 9OLT 2 — = 6 F2 al)
spue[st ивәившәџрәр
Vol 11 08 F2 OOS LT F2 = OOOT F2 ТӨС 1воѕедереу
VL € 18 9661 ОТ 666 897 рив[294 мәм
9°89 IT “OL 98 8811 SI L9G 9'9[ емен
5рир[в1 әјвләйшәј 10 perdoj,
CLO 6 cr F2 0062 == 0081 SL вшвивд
L991 8 0006 "BJ = 1181 YS BOTY 81509
sa1do1] 19M о] ISIOJA
9'6c L SL TD 0008 F2 10
Cel 9 89 0€9€ Og F2 COV OLE biſpasnv шәјәмпос
Vol © VES 9? LESS = = SSI c
O'S Y CGE ВӘ COLE — [6S TOI 1 91149 189997)
Vel € EL Over G9 908 VEE 22ut^01d әпѕцор erojen
£001 1°89 0£06 rol 066 06 eaujy YINdg ‘uoIsay ade?)
әјешцә URVURLA}IPeul JO ввәлү
gel c TIE OOS 9T Uc Olcc €8cI Mod
97 I €708 2960 [ША OFT? T29c eoujy шәщшпос
searP [eJuaunuor),
UY OT 32u219J29M Y satoeds ж в1әиәЗ Jo ¿MY «0T
/sotoadg 2muapuo [F10], әїшәриә 19QUINN] Bale
satoeds в1әиә‹) [Bossy
(2661) 122200 Y [EPIN “ZI (9661) 1° 19 z
`6 їсшшоә "srod) шпАрлсу y јәшшең ^g :(z66[) 19ddog ^7 :(0761) P94 `9 (7661) ѕәләтлвә) y ooy uey С
YS “LL 40661) TP 19 19U (1661) 192€ “OT (2861) AAV.
“(P661) [E 19 олошү uey ^t (8/6[) родәхү Y USAPY ^e ‘(wwo
ләй) rgoonaez *(e66[) nqoonaez y oyeg `g (wuwo *s1ad) Jay әр 77) `1 ‘SQOUIIIJIY "ppro« әці jo ѕиог8әл p3122[as ut sjue[d re[nose4 әлцви јо wsiwəpuə jo uosueduio") сб әү4еј,
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
1978). under-
scores the peculiarities of the Cape flora (Table 9).
47.7% endemic (Raven & Axelrod,
Such levels of endemism are usually associated
with islands that have been isolated for long periods
of geologic time, or with areas that have very sharp
boundaries limiting direct plant migration. In a bi-
ological sense, the Cape Flora Region is virtually
an island, not surrounded by ocean, but by a zone
of dry climate or sharply different soils, or of sea-
sonal rainfall distribution. The flora of southwestern
Australia shares with the Cape Region an unusually
high endemicity for a local continental flora (Beard,
1970: Hopper, 1992), and so does that of southern
Africa as a whole (Table 9). Why the latter region
should have such a high level of endemism is not
at all clear.
The high level of diversity and local endemism
of the CER is starkly emphasized by comparison of
the phytogeographic centers of the region with other
The SW
4000 species and over
and NW Centers each have over
31% or 26% endemism, re-
spectively, compared with about 2400 species and
areas.
an estimated 22-23% endemism (depending on the
geographical definition of the region) for the entire
mediterranean flora of Chile (Table 9), an area over
five times greater than either of these. phytogeo-
graphic centers of the Cape flora. Likewise, impor-
tant centers of local endemism (hot spots) within
the Mediterranean Basin, including the islands of
Sicily, Sardinia, and Crete, or the Peloponnese Pen-
insula, all of more or less comparable size to the
NW or SW Centers, have approximately half or less
than half the number of species and between 6%
and 12.5% endemism (Médail & Quézel, 1997).
ALPHA DIVERSITY
Aspects of plant species diversity have been ad-
dressed on several levels. and it has been shown
that at the local level selected areas within the
unusual on a world scale
1992; Goldblatt, 1997) and are less
species-rich than many areas sampled in the New
1988a.
1988b). Patterns of alpha diversity (the number of
Cape Region are not
(Cowling et al.,
and Old World lowland tropics (Gentry,
species in a homogeneous community, e.g., Cowling
et al., 1992) in a range of vegetation types in the
Cape Region, including fynbos, renosterveld, forest
thicket, and evergreen forest, are surprising. Fyn-
bos sites (with seasonal species not included in
species counts) have a mean alpha diversity of 68
species рег 1000 nv, with 121 the highest number
of species recorded at any site (Cowling et al.,
1992).
and data indicate that diversity is higher in the west
Fynbos diversity is by no means uniform,
than the south and in sites of intermediate produc-
tivity than in either more mesic or drier sites (Bond,
1983).
Non-fynbos sites have been less well studied.
Figures in the literature for renosterveld include
means of 66 (Tilman et al., 1983) and 84 (Cowling
1992) species per 1000 nr. Forest thicket
sites have a mean of 59 species per 1000 nr, forest
et al..
sites have ranges of 44 to 52 species, and succulent
karoo shrublands a mean of 43 species in the same
area (Tilman et al., 1983).
Comparisons made by Cowling et al. (1992) in-
dicate that California chaparral communities have
alpha diversity levels around 34 species per 1000
m? but the more comparable southwestern Austra-
lian region has an average of 69 species per 1000
m? in heathland (kwongan), a vegetation type sim-
ilar to fynbos. While these figures are consistent
with higher total species richness in the CFR and
southwestern Australia compared with the Califor-
nia Floristic Province, they do not explain the com-
parable regional diversity of the CFR and the low-
land wet tropics. In the wet tropics mean alpha
diversity of trees (including woody lianas) alone has
been found to range from 129 species (Africa) to
140 species (Neotropics), to 193 species (Asia) per
1000 m? (Gentry, 1988b), about twice the alpha di-
versity found in mediterranean communities on nu-
trient-poor soils. Inclusion of epiphytes and other
herbaceous plants raises alpha diversity in tropical
sites. Gentry and Dodson (1987) have shown that a
major component of the plant species diversity in
tropical forests actually lies with the epiphytes.
Similarly the inclusion of seasonal geophytes would
increase the figures for the Cape flora. and until
more Comprehensive surveys are made any com-
parisons can only be tentative.
BETA AND GAMMA DIVERSITY
Beta diversity (species turnover along a habitat
Cowling et al., 1992) is
Cowling (1990),
for example, has reported nearly complete replace-
or environmental gradient,
relatively high in the Cape Region.
ment in sites along the Agulhas plain that differed
-
=
| soil features but were climatically and topo-
— similar. Differences in composition be-
tween Communities on sandstone versus clay soils
are so nearly complete that the plants on these two
soils are treated as belonging to different vegetation
types.
Gamma diversity (species turnover in equivalent
eradients, also called
1992) is likewise ex-
tremely high in the Cape Region, and is reflected
habitats along geographic
delta diversity, Cowling et al..
296
Annals of the
Missouri Botanical Garden
in the high levels of regional endemism. Species
replacement values of 46-70% have been reported
by Kruger and Taylor (1979) for sites 25 km apart,
and Linder (1985) has suggested that geographic
replacement may account for 3046 of the differenc-
es in species composition along geographic gradi-
ents in similar habitats. These figures are, however,
lower than some estimates for lowland Neotropical
sites (B. Boyle in Goldblatt, 1997).
REASONS FOR CAPE FLORISTIC DIVERSITY
Richer in species than any other temperate flora
and most tropical ones occupying comparable phys-
ical area (Table 9), the Cape flora is also highly
distinctive. One of five regions in the world with a
mediterranean climate, the Cape has substantially
more species than do either California or central
Chile, which are substantially larger in land area.
Although southwestern Australia has a flora that
may have about the same number of species as the
Cape Region (Table 9), it is at least three times
larger in area. Only the Mediterranean Basin, ap-
proximately 25 times larger in area, has a flora that
is larger than the Cape flora, with about 2.5 times
as many species. A formal comparison of these pat-
terns is provided by Cowling et al. (1996). The rea-
sons for the substantially higher species diversity
of the Cape Region compared with these floras are
several, and include a range of factors, both phys-
ical and historic.
A diversity of soils, a rugged landscape, and ex-
tremely variable and complex rainfall patterns have
combined in the Cape Region to produce a mosaic
of sharply different habitats that lie in close prox-
imity to one another in a pattern repeated across
its entire area. This high physical heterogeneity, al-
though striking, is not unique to the region, and
may even be greater in other regions. For example,
the California Floristic Province has a wide range
of soils, including serpentine substrates not present
in the Cape Region, diverse climates, a rugged to-
pography with higher mountains than those in the
Cape, plus a wider latitudinal extension. Likewise,
the Mediterranean Basin, orders of magnitude larg-
er than the CFR, has a wide diversity of soils and
a rugged landscape, with the mountains also higher
than those found in the CFR. Both areas are often
cited as being species-rich compared to neighbor-
ing regions, yet both are substantially poorer in
species than the Cape Region, California in abso-
lute number, and the Mediterranean area in sub-
stantially less alpha diversity per unit area. South-
which has a flora
western Australia, in contrast,
approaching that of the Cape in size, noticeably
lacks the rugged topography of other Mediterranean
regions, although it exhibits some of the ecological
features of the Cape Region. Physical heterogeneity
alone cannot therefore account for the richness of
the Cape flora, and edaphic factors as well as his-
torical biogeography may be more significant.
Of the five mediterranean regions of the world
only the Cape and southwestern Australia have
soils that include large areas of nutrient-poor
quartzitic sands, and at least part of the explanation
for the higher species numbers here has been
thought to relate directly to the particular effects of
this substrate on plant life. High levels of local di-
versity have been considered characteristic of nu-
trient-poor soils (Tilman, 1982, 1983) and if this is
correct, then the mere existence of large areas of
such soils should account for the comparatively
high alpha diversity of heathland vegetation in both
South Africa and southwestern Australia compared
with that in California or central Chile unless high
levels of local and regional richness are not cou-
pled. This hypothesis is, however, not supported
either in the Cape or southwestern Australia, where
alpha diversity levels on nutrient-poor and nutri-
ent-intermediate soils appear to differ only mini-
mally (Goldblatt, 1997). Although nutrient-poor
soils may not support a flora significantly richer
than that occurring on soils of other nutrient status
in the Cape, they are able to maintain particularly
high beta diversity in the associated fynbos vege-
tation, both on geographical gradients and on
slightly different soils under the same climatic con-
ditions. Comparable data for nutrient-intermediate
and nutrient-rich sites are not yet available. Nutri-
ent-poor soils in mediterranean climate zones have
a higher proportion of reseeding versus resprouting
shrubs (Wisheu et al., 2000), and these authors ar-
gue that this directly contributes to higher diversity
because the high frequency of fires that destroy re-
seeders would lead to their shorter generation times
and thus higher speciation rates. The high frequen-
cy of fire in areas of low-nutrient soils is thus an-
other aspect that may contribute to diversity in the
Cape Region. Fire is also significant in southwest-
ern Australia and California but not in the medi-
terranean zone of Chile (Kalin Arroyo et al., 1994).
The combination of edaphic and topographic di-
versity, steep local climatic gradients, peculiar nu-
trient-poor soils, and frequent fires, although un-
doubtedly important in promoting species diversity
in the Cape Region, is still inadequate to explain
the presence of the unusually rich flora in the Cape
Region, particularly when compared to southwest-
ern Australia. A notable and perhaps crucial dif-
ference between the Cape and other areas of med-
Volume 89, Number 2
2002
Goldblatt & Manning 297
Plant Diversity
iterranean climate, possibly excluding southwestern
lies in their Pliocene-Pleistocene cli-
Available data indicate a history for
Australia,
matic
southern Africa very different from that experi-
enced in the Northern Hemisphere and to a lesser
extent Chile (Villagrán. 1994).
cold and aridity alternating with warm wet phases
history.
Cycles of extreme
made these areas largely uninhabitable by anything
resembling their current floras, elements of which
either became extinct or were locally restricted to
sites of relatively mild climate. A similarly dynamic
history for the Cape flora has been hypothesized in
which Pleistocene glacial cycles caused a north-
ward shift in the winter-rainfall zone, which in turn
caused a northward expansion of the flora during
the pluvial periods but local extinction and its re-
striction to refugia during dry periods (Axelrod &
Raven. 1978). The limited evidence available for
the Cape Region, however, indicates a more mod-
erate climate without changes of such cataclysmic
dimensions.
modest. shifts in the flora (Meadows & Sugden,
199]
sitive Cedarberg mountains.
Pollen cores reflect. comparatively
. even in the semi-arid and ecologically sen-
Changes there might
—
be expected to have been more severely influenced
by the drier and colder climatic conditions that are
postulated to have occurred during glacial periods.
when belts of vegetation adapted to mediterranean
climates contracted away from the dry zones that
lay toward the equator. The absence of any evi-
dence of major changes in the vegetation of the
Cape Region during the Pleistocene suggests that
the glacial cycles did not have the catastrophic ef-
fects on plant life in southern Africa that they did
in the Northern Hemisphere or Chile. In the CFR
relatively drier and wetter cycles may simply have
induced changes in the local composition of vege-
tation, perhaps causing limited extinction, which in
turn created opportunities for speciation.
The unusually high species richness of the Cape
flora is thus. in all likelihood, a consequence of
sustained climatic stability and reliability, enabling
a more or less uninterrupted evolution and diver-
sification of the flora to occur in a region of high
physical complexity. The history of this evolution
can be traced to some extent by considering the
Although
there are few such studies for plant groups centered
modes of speciation evident in the flora.
in the Cape flora, the available evidence suggests
that parapatric speciation linked to substrate or mi-
croclimatic differences is an important mode of
speciation in some families (Linder & Vlok. 1991:
Cowling & Holmes, 1992b: Goldblatt & Manning.
1996). Vicariant species exhibit differences in ecol-
ogy such as edaphic, microclimatic, seasonal, or
pollination characteristics. For example. vicariant
species of Rhodocoma iBastionae eae). a genus re-
stricted to nutrient-poor sandstone soils in montane
habitats, favor different habitats and are not signif-
icantly isolated geographically (Linder & Vlok.
1991). Parapatric speciation also appears to have
geographic isolation in
1992b)
and in the genus Lapeirousia (Iridaceae) in the CFR
been more important than
Agulhas Plain shrubs (Cowling € Holmes.
5 8
and adjacent parts of the southern African west
coast (Goldblatt & Manning, 1996). High levels of
both beta and gamma diversity likewise support the
hypothesis that microgeographic speciation has
played a major role in speciation in the Cape flora.
Nearly adjacent habitats with the same climatic and
topographical conditions, which differ only in their
substrates (coarse sand or fine sand or limestone).
can support plant communities that differ radically
in their species composition while still being broad-
ly similar in family and generic composition (Cowl-
ing & Holmes, 1992a)
Parapatric or microgeographic speciation may
actually be the rule not the exception in plants
1993), and is favored by reduced gene flow
f
contrasting substrates that characterizes the CFR
(Levin.
across strong selection differentials. The mosaic
appears to provide such a strong selective differ-
ential. Although the Cape flora may not differ from
other floras in mean pollen dispersability (Linder,
1985).
have low seed dispersal distances. The majority of
many of its most characteristic elements
species in the Cape Region show no evident ad-
aptations for seed dispersal and are regarded as
passively dispersed, with estimated seed dispersal
distances under 5 m (and most likely much less
than this) (Linder, 1985). Dispersal in most Aizo-
aceae is by rain drops falling on hydrochastie cap-
sules, and this mechanism, although an active one.
results in very small dispersal distances (Desmet &
Cowling, 1999).
disproportionately well represented in both the
Cape Region (Bond & Slingsby, 1983) and Austra-
lia (Berg, 1975). Some 1000 Cape species, notably
Proteaceae. Restionaceae.
Active seed dispersal by ants is
in the families Fabaceae.,
Rhamnaceae. and Rutaceae, produce seeds with
lipid bodies (elaiosomes) that are attractive to ants,
and an undetermined additional number are trans-
ported to underground nests by harvester ants. In
vegelation types prone frequent fires, such as
fynbos, the burial of seeds is a valuable adaptation
1992p).
distances for ant-dispersed seeds are also short. up
6 m (Linder, 1985).
strategies are relatively restricted in their occur-
(Cowling & Holmes, However, dispersal
More effective dispersal
rence. Wind-dispersal is characteristic of many As-
298
Annals of th
Missouri Botanical Garden
teraceae and Orchidaceae (which have compara-
tively low levels of endemism in the Cape flora),
while long-distance dispersal involving flying ver-
tebrates (birds and bats) is least common, and is
especially rare in plants on nutrient-poor sub-
strates. There is an assumption that plants on such
soils cannot afford to allocate resources to protein-
rich berry or drupaceous fruits (Bond & Slingsby,
1983). The low frequency of fruits with burs and
spines, adapted for exochory, suggests that dispers-
al by non-flying vertebrates has long been unim-
portant in the flora, perhaps because the fauna has
historically been a small one in terms of numbers
of species and individuals. This may be due largely
to the unpalatability, low nutrient status, and low
productivity of the flora as a whole.
Indirect evidence of the importance of reduced
gene flow distances in stimulating local species di-
versity Comes from a comparison of the number of
species and their level of endemism between taxa
with widely dispersed seeds and those with reduced
dispersal distances. Genera with fleshy diaspores
or those that are well adapted for wind dispersal
tend to have wide ranges, few species per genus,
and low levels of local endemism. Compare the ber-
ry-fruited Nylandtia (Polygalaceae). which has two
species, with its relative Muraltia, which has dry
fruits and over 100 species, most with narrow rang-
Similarly, Chasmanthe (Iridaceae), which has
fleshy or deceptive (brightly colored) seeds, has two
widespread species and one localized one, whereas
its close relative Tritonia, which has dry seeds, has
es.
16 species in the flora, most of them with narrow
ranges. Another striking example is Chrysanthe-
moides (Asteraceae), which has seeds enclosed in
a fleshy pericarp. The two species of the genus ex-
tend throughout the Cape Region and one far be-
yond it into tropical Africa. The numerous species
of the closely related genera Osteospermum and
Tripteris have dry seeds and mostly have smaller
geographic ranges. This comparison is also instruc-
tive at the family level. Low seed dispersability is
typical of many of the larger and most characteristic
families in the flora with high ratios of species to
genus and high levels of local endemism. Erica-
ceae, Iridaceae, and Fabaceae, which largely lack
highly developed mechanisms for long-distance
seed dispersal, have ratios of above 20 species per
genus and higher than 80% endemism at the spe-
cies level. Levels of local and regional endemism
for Asteraceae (63% endemic species) and Orchi-
daceae (60.8%), mostly with wind-dispersed seeds,
and Poaceae (38.6%) and Anacardiaceae (32.1%),
many with fruits adapted to wind and/or animal dis-
persal, show levels of local and regional endemism
below the mean for the entire flora (68.7%).
The massive speciation in the Cape flora is, we
suggest, most likely explained by a model of local
speciation in the absence of catastrophic climatic
or topographic perturbations. It appears that a rel-
atively stable climate prevailed in the Cape during
the Pleistocene and that local parapatric speciation
across steep environmental gradients may account
for a considerable proportion of the speciation
events that occurred in the CFR. Because of this
relative stability it seems reasonable to postulate
that extinction rates in the main vegetation zones,
fynbos, renosterveld, and succulent shrubland,
were low and more than compensated for by local
speciation events. The nutrient-poor soils scattered
in a mosaic across the CFR must have stimulated
local speciation rates because of the characteristic
low vagility of the seeds in the great majority of the
plants adapted to these soils. Likewise, the low va-
gility of many of the species in Succulent Karoo,
although likely a result of different selective forces,
has the same consequences, with high levels of lo-
cal speciation, and thus high levels of diversity
across geographic, environmental, and edaphic gra-
dients (e.g.. Cowling € Holmes, 1992}; Cowling et
al., 1998; Desmet € Cowling, 1999).
The unusually high diversity of the Cape flora is
matched by its extraordinary composition of fami-
lies and genera. Instead of a balanced flora with
relatively small numbers of species per genus there
has been massive local radiation in a series of un-
related genera. This is so pronounced that almost
22% of the total species in the Cape Region fall in
just 10 genera, while the 20 largest genera account
for over 30% of the total species (Table 3). Typical
examples of these genera are Erica (over 650 spp.).
Aspalathus (272 spp.). Agathosma (143 spp.), Phy-
lica (133 spp.). and Cliffortia (114 spp.). Signifi-
cantly, none of the genera that display a pattern of
massive local radiation in the Cape Region are en-
demic there, but rather they extend northward into
Namaqualand, eastward into southern KwaZulu-
Natal, or even further to the northeast into tropical
Africa. They are, however, primarily restricted to
nutrient-poor soils wherever they occur.
In sharp contrast to these examples are the en-
demic families of the Cape flora, which are without
exception depauperate in species, although they
may contain several genera. These families are typ-
ically restricted to montane habitats in acidic sand-
stone soils, and many of their constituent species
are highly local endemic plants of particular moun-
tain chains or peaks. They display the character-
istics of paleoendemism and probably represent el-
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
ements of a previously more widespread southern
temperate flora adapted to nutrient-poor soils in a
summer-rainfall regime. It is probably no coinci-
dence that most of these species flower in summer,
and are thus out of phase with the spring flowering
peak of the flora. With the development of a winter-
rainfall climate in the Pliocene (e.g... Coetzee,
1993), it is reasonable to infer that these pre-Cape
flora elements were gradually restricted to mesic
sites in which some moisture was present over the
summer months. Concomitantly it appears that oth-
er elements of the flora were able to radiate into
emerging niches, thereby establishing the huge
neo-endemic element of the flora. The highly scler-
ophyllous or microphyllous habit developed by taxa
adapted to nutrient-poor substrates can thus be re-
garded as an important pre-adaptation to the med-
iterranean climate. The hi
the mediterranean climate, as well as its regularity,
would also have favored families and genera with
a geophytic habit,
for example, Amaryllidaceae,
Hyacinthaceae, Iridaceae, and Oxalidaceae.
The rapid and extensive radiation of plant taxa
in the CFR must have been favored by both the
emergence of new habitats through climatic change
and the exposure of the coastal plain as sea levels
fell at times during the Pleistocene (Coetzee, 1993),
At the
same time the flora was increasingly isolated by the
as well as by the stability of the climate.
winter wet and summer dry climate regime from
recruitment from the summer-rainfall-adapted flora
of adjacent parts of southern Africa, which largely
lack the low- or nutrient-intermediate soils so char-
acteristic of the Region.
of the (
diversity of flower form
final characteristic Cape flora is the
great, and often extreme,
that is a feature of many of the genera. This is
linked to a diversity of pollination strategies, many
of which are poorly exploited outside the region. In
particular, pollination by sunbirds, long-proboscid
flies, hopliine beetles, rodents, and the butterfly
Aeropetes аге more extensively Haw ni in the
a (Manning &
1998; Goldblatt
All these strategies favor dif-
Cape flora than. elsewhere in
Goldblatt. 1996; Goldblatt et T
& Manning. 1999).
ferently shaped. large. brightly colored flowers. Pol-
linator diversity is likely to be part of the expla-
nation for the high species diversity in the CFR,
particularly for certain families in the flora. Both
Iridaceae and Ericaceae, for example. have adopted
a range of pollination strategies not evident or only
weakly expressed elsewhere across their range (Vo-
gel. 1954: Goldblatt & Manning. 1996, 1998;
hardt & Goldblatt, 2000), and are often particularly
striking when in bloom.
; Bern-
In Gladiolus, a prime ex-
ughly seasonal nature of
ample of a genus with diverse pollination systems
and represented in the CFR by 106 species. the
genus exhibits no less than 27 shifts in pollination
system in southern Africa, most of these in the CER
(Goldblatt et al., 2001), and repeatedly evolved ad-
aptations for long-proboscid Ну, sunbird. moth, and
butterfly pollination, which must have played a ma-
jor role in its radiation. The short season favorable
for both plant and insect growth is probably the
overriding factor responsible for the diversification
of pollination strategies and, more than anything
has made the
else. Cape flora so extremely ap-
pealing to human sensibilities.
CONSERVATION
Like other parts of the world with species-rich
floras and an expanding human population, there
are varied and serious threats to the Cape flora.
Expanding agricultural activity for growing food or
to supply other human needs has transformed low-
and areas near population centers to the extent that
little or
1992). Moreover, the peculiarly local distribution
no native plants remain locally (Rebelo,
of many Cape plants (high beta diversity) means
that a higher proportion is imminently threatened
with extinction than would be the case with wide-
spread species. The number of species known to be
lost forever is relatively low, but the number of spe-
cies represented by single, reduced, and diminish-
Taylor (1978
species in the
ing populations is alarmingly high. '
estimated that some 500 endemic
CFR were threatened,
were known to be extinct (0.07%).
while at least 60 species
Red Data Book
numbers for southern Africa in 1985 include 1320
Cape plants (14.67% of the total) (Hall & Veldhuis.
1985). Criteria for classification of threatened spe-
cies vary: Rebelo (1992) estimated 218 threatened
species (extinct, endangered, or vulnerable) in the
CFR (ca. 4.8%), slightly less than half the figure
provided by Taylor, but nevertheless, considerable.
In 2001 these figures are certain to be higher as
^
result of the expanding human population. de-
spite ever more sophisticated conservation activity.
\ peculiar threat to the Cape flora is the spread of
alien vegetation, largely woody species of Austra-
lian Acacia and Hakea and European Pinus. These
species rapidly invade montane and lowland areas
not currently suitable for agriculture, thus com-
pounding the threats of agricultural activity and ur-
ban growth. The threat posed by this alien vege-
—
ation is being vigorously countered by biological
control methods and manual removal, but the prob-
lem is immense and requires constant monitoring.
Kruger (1977) estimated that 60% of fynbos had
300
Annals of the
Missouri Botanical Garden
been replaced by alien vegetation, agriculture, or
urban development. ‘Twenty-five years later, this fig-
ure is bound to have increased.
Concrete examples may better illustrate the sit-
uation. Rebelo (1992) estimated that of 306 endem-
ic species of Proteaceae 65 were threatened (de-
vulnerable) (ca.
21%), while 131 were treated as Red Data species.
The pattern in other families is similar, although
fined as extinct, endangered, or
slightly less serious, with 10.5% of endemic Iri-
daceae, 996 of endemic Rutaceae, ca. 396 of en-
demic Ericaceae, ca. 2% each of endemic Astera-
sae threatened (Rebelo, 1992). The
number of species of each of these families includ-
ceae and Fabac
ed in the Red Data Book is substantially higher
(Hall & Veldhuis, 1985
A detailed analysis of reserves and other
—
con-
servation areas in the CFR is presented by Rebelo
(1992).
substantial areas for preservation of diverse vege-
Recent conservation efforts have secured
tation types in the CFR and a framework for a con-
servation plan to establish reserves to cover por-
tions of all vegetation types has been developed
2001).
Although such reserves cannot include populations
(Cowling & Heijnis, 2001; Cowling et al.,
of all endangered species because of their erratic
distributions, future loss of distinctive vegetation
types will be limited and preservation of substantial
tracts of each will in the future be enhanced.
Literature Cited
еа Phylogeny — (APG). 1998. An ordinal
sification for the families of flowering plants. Ann.
Missouri Bot. Gard. ч 531—553.
“Н. & B. К de Wet (editors). 1993. Plants of
Africa. Names and Distribution. Mem. Bot.
Surv. S. Africa 62. National Botanical Institute, Preto-
ria.
Axelrod, D. I. & P. H. Raven.
Tertiary vegetation history of / .
A. Werger — ta and Ecology of
Southem Africa. W. Junk, The Hague.
Bayer, M. B. 1984. m Cape flora and the karoo—A win-
ter — اا versus a fynbos biome. Veld & Flora
70(1):
Beard, — pro^ (editor). An Annotated Checklist of the
Plants of bó stern Australia. Kings Park Board, Perth.
Berg, R. Y. 1975. Myrmecochorous «кү in — and
their dis pera by ants. Austral. J. Bot. 23: 475-508.
Bernhardt, P. & P. Goldblatt. 2000. The aeui of pol-
lination mec inns in the — of southern Afri-
ave ые Cretaceous and
1 Pp. 77-130 in М
ca. Pp. 301- 1 K. Wilson & P. Weston (editors).
Systematics e Bios of the — ‘ots. Royal Botanic
ardens, Sydney, Australia
devis of the ( в : lora:
African Bot., Suppl. 13
On — dive rsity and the ric [ав of
the Cape боа: А study in ugs rn. Cape — P
— 13 in F J. Kruger, D Mitchell & J.
mate Ecosystems. Springer-Verlag, Berlin.
———— & E.
Jarvis (editors), Mediterranean-Type oe ms. The
Role al pias nts. Springer- Verlag. Be
— —— & P. Slingsby. Seed need by ants in
— a the Cape и and its evolutionary
implications. S. African J. Sci. 79: 231-233.
Brako, L. & J. E Zaruce hi. 1993. Catalogue of the Flow-
i eru. Monogr. Syst.
and e of
. Missouri Er Gard. ‹
B B., R. job (& J. Sweere.
1994. rbcL sequences support in exclusion p Retzia,
Desfontainia, and s fox from the Gentianales. Pl.
Syst. Evol. 190: 2 123 0.
Chase, M. W., D. E. a. P. S. Soltis, P. J. Rudall, M.
К. jet W. H. p 5. Sullivan, J. е h, М. Molvray,
P. J. Kores, T. J. Givnish, K. J. S a & J. C. Pires.
2000. Higher- but са s of he monocotyledons:
An assessment of current cigar and a new clas-
sification. Pp. 3-16 in К. Wilson & P. Weston (editors),
Systematics and Biology of the Monocots. Royal Botanic
Gardens, Sydney.
Coetzee, J. A. 1993. African flora since the terminal Ju-
rassic. Pp. 37-61 in P. Goldblatt (editor), Biological Re-
ae Between’ Africa and South America. Yale
Univ
=ч
ress, New Hav
игы
— 7
. Rogers. 1982. Palynologic val and —
Mu nce for the Miocene — nvironment in the
danha — — Africa). Palaeogeogr. —
Palaeoecol. 71-85.
& — 1985. The phytogeographic signifi-
cance of some extinct Gondwana pollen types from the
Tertiary of the و — (South Afric а). Ann.
Missouri Bot. Gard. 71: 1088-10
— —— & lL. Pra (ues 1984. Pus evidence for the
occurrence of Casuarinaceae and Myrica in the Tertiary
of South Afric "ul. 23: 23—
Conran, Chase & P. I Rudall. 1997. Two
new monoc йу айо families: Anemarrhenaceae and
ie — ‘eae (Lilianae: Asparagales). Kew Bull. 52: 995—
Grana
~
Cooling, R. M. 1987.
speciation in Gondwanan shrublands. S. African J. Sci.
83: 106-112.
1990. Diversity compone nts in a spec ies-rich
area of the Cape Floristic Re холоп. J. Veg. Sci. 1: 699-
Fire and its role in coexistence and
7
1992.
Press, с Tow
& ( Heli ijnis. 2001. The identification of broad
habitat units as biodiversity entities for systematic con-
se — — in the Cape floristic region. S. Afri-
: 15
сап E
bei Ecology of Fynbos. Oxford Univ.
— 1992a. Flora and vegetation. Pp.
23-01 2 M. Cowling (editor), The Ecology of Fyn-
bos. Oxford Univ. - 'ess, Cape Town
& )2b. Endemism айд speciation in a
—
lowland flora s hei Cape Floristic Region. Biol.
Linn. Soc. 47: 307—
& D. J. Me Donald,
plant dede i
17
(е Ae
1999. Local endemism and
^W Ru ndel, G. Montenegro & F. J
bandi Degradation in Mediterranean-Cli-
—
T. E Witkowski. 1994. Convergence and
non-convergence of plant traits in climatically and
— ‘ally matched sites in saree an Australia
and S ree Africa. Austral. J. Ecol. 19: 220-232.
.H olmes & A. С. Ки
1992. Plant di-
Volume 89, Number 2
2002
Goldblatt & Manning
Plant Diversity
versity and endemism. Pp. 62-112 in R. M. Cowling
— — Ecology of Fynbos. Oxford Univ. Press,
Cape p /
——., P. ndell, B. B. Lamont, M. Kalin Arroyo &
M. St. 1996. Plar
climate — s. p — Evol. 362—
————, P. W. Rundell, P. C. Desmet ^ ^ Tea 1998,
шоо d T -scale plant diversity in
southern African aid lands: Subcontinental and global
Diversity & Distributions 4: си.
—— ‚ Cartwright, J. E. Parki Allsop.
1999. Fossil wood charcoal assemblages ов. ed
ie Cave, South. Africa: Implications for Late Quater-
ary vegetation and climates i in the winter-rainfall fyn-
x Biogeogr. 26: 307—378
= ssey, A. T. Lombard, C. E. Heijnis, D.
Mie еа & N. Cole. 2001. Framework for a Con-
servation Plan for the Cape Floristic Region. Institute
: Plant Conservation Report 9902 submitted to WWF:
it r in Сва
c отрап isons.
yo |
hos Бае:
p nm W. D. 1987. Flora of Panama. Checklist and In-
dex. Missouri Ке
19 ee ology. Pp. 58- 66 ‚ Day,
. Sici. À N. Louw & M. L. Jarman (editors),
A Preliminary — S. Africa
Sci. foar kep 40. CSIR, Pretor
P. G. & R. M. Cowling. 1999. Biodiversity. hab-
itat and range-size — of a flora from a winter-rain-
fall desert in north-western Namaqualand, South Africa.
23-33.
D
Garden, St. Louis
at.
Desmet,
Esler, K. J. 1998. Unusual geophytes of the succulent ka-
roo. з & Flora 84: 6-9
› W. Rundell & P. ete. 1998. Bioge ography
of тк leaved geophytes in semi- i South Africa:
Hypotheses | on functionality. Pl. Ecol. : 105-120.
Fenner. M., W. G. Lee & J. Barstow ane 1997,
comparative slt idy of the alistribulion of genus size in
twenty angiosperm floras. Biol. J. Linn. Soc. 62: 225-
287.
—
Gentry H. 1988a. Tree species ric — of uppe r —
azonian species. Proc. Natl. Acad. Sci. U.S.A. ¢ 56-
159
1988b. Changes plant community diversity
and floristic composition on environmental = geo-
graphic val gradients. Ann. Missouri Bot. Gard. 75: 1-34.
— & C. Dodson. 1987. Diversity and ее — у
ШП у: vascular epiphytes. Ann. Missouri Bot.
ard. 233.
быды, А oe
lts characteristics, ir a and origins.
Au Bot. Gard. 65: 369—
1979. Miscellaneous кыны! counts in an-
An analysis of the flora of southern
P ncluding new family and generic. re-
Gard. 66: 865-86
giosperms,
Missouri Bot.
Floristic —
sity Conservation 6:
cords. Ann.
1997.
South — a. 359-311.
& J.C. ›. Phylogeny and speciation
in Layers — Laer — eae subfam-
ily BE deae). Pu Bot. (
Gladiolus in eben
Fe e Press, Cape Тип п.
& ————. 1999. The long-proboscid fly pollina-
Ga ud. € 516—301.
Africa.
: ion — m in Gladiolus | (Iridaceae). Ann. Missouri Bot
ard.
Es — 2000. Cape Plants. National Botanical
Institute, Cape Town.
— ———, К. Dahlgren, B. J. Nielsen & J. P. Rourke. 1979.
in the Cane Flora of
Further notes of Retziaceae: Chemical contents and af-
finities. Ann. Missouri Bot. Gard. 66: 545-550.
‚ P. Bernhardt & J. C. Manning. 1998. Pollination
of ita aa geophytes by monkey beetles (Scarabaeidae:
Rutelinae: Hopliini) in southern. Africa. Ann. Missouri
Bot. Gard. 85: 215-230.
. €. Manning & P. Bernhardt. 2001. d
of pollination systems in Gladiolus (Iridace:
de — in southern Africa. Ann. Missouri Bab
34.
: Crocoi-
Gard. 88:
б R 1974. The Geography E the Flowering Plants,
ed. 4. Longmans Green, — or
Groves, R. H., J. S. Bea H. E Deac on, J. J. Lam-
N.
Specht & W. D.
origins and character-
A
Rabinovite de Vin,
Introduction: e
brechts, A.
Stock. 1983.
istics of mediterranean ecosystems. Pp.
Day (editor), Mine ral Nutrients in Mediterranean Eoo-
can Nat. Sci. Progr. Rep. 71. CSIR -
-
1 in J.
syste ms. S. Afr
Ha & H. A. Veldhuis. 1985. South Mrican Pe
Data As — a karoo biomes
ep. SIR, Pretoria.
m
>:
Au
te
m s A Record of Past
Life. South hacen Museum, Cape Town.
Hopper, S. D. 1992. Patterns of plant diversity at the pop-
ulation and species levels in south- west
Part E báo subdivision. Vegetatio 97: 21—
38.
————,‚ 1997.
arid regions.
Floristic biodiversity and history of African
Biodiversity & Conse — 6: 195-514.
The mediterra-
What do we
Notic. Biol.
Kalin Arroyo, M. T. & L. Cavieres
nean type-climate flora of — nc
know P how e
48-5
an we assure ils protection,
. С. Marticorena € M. Munoz-Schick.
1994. с onvergence in the Mediterranean floras in Cen-
tral Chile end California: Insights from comparative bi-
ology. Pp. 43-88 in M. T. Kalin Arroyo, P. H. Zedler &
M. D. Fox (editors), Ecology and Biogeography of Med-
California Austra-
lia. w Stud. 108. Springer-Verlag, New Yo
02. The Geology of the Earth. Oliver & Boyd.
‚ 1977. The ecology of early man in southern
Africa. E тепсе 197: 115-126.
Kruger, F. J. 1977.1
Toward a strategy for conservation. S. African J. Sci. 73:
81-85.
& H. C. Taylor. 1979. Plant species diversity in
Le ologic al reserves in the C ape fynbos:
TO
В fynbos: Gamma and delta diversity. Vegetatio 41:
8593.
Lebrun, J.-P. & A. L. Stork. 1997.
tes à Fleurs d? Nie Trogi ale. 5: Conse
din Botanique de
Levin, D. A. 1993. Loc e spec — in — The rule
d O Syst. Bot. 18: 197
85. Gene flow, spec ialion pem species di-
Enumeration des Plan-
ervatoire el Jar-
a Ville de Genéve, Geneva.
L inder,
ve rsity pev in a species-rich area: The Cape flora.
Pp. 53-57 in . Vrba on Species and Specia-
tion. Transvaal A Monogr. Transvaal Museum,
Pretoria.
& J. H. J. Vlok. 1991. The morphology. taxonomy
302
Annals of the
Missouri Botanical Garden
„Р
—
and — of Rhodocoma (Restionaceae Syst.
Evol. 175: 139-1 —
Louw, ^ B. & A. G. Rebelo. 1996. Vegetation of South
Africa, Lesotho pe Swaziland. De ‘partment of Environ-
mental Affairs oa Tourism, Pretoria
Maggs, G.L., P. ¢ n & Н. Kolberg. 1998. Plant species
richness, uU M: and genetic resources in Namibia.
Biodiversity a Conservation 7: 435—440.
Manning, J. €. & P. Goldblatt. 1996. The Prosoeca pr
ingueyi (Dipte ra: Nemestrinidae) pollination syndrom
in southern Africa: Long- — sd flies and their tu epa
lowers. Ann. Missouri Bot. Gard. 83: 67-86
Meadows, M. E. & J. M Sue n. 1991. A vegetation his-
€ of и last 14 00 ) years on baie — south-
Cape Province. 5. African J. Sci. 87: 3
Médail, E & P. Quézel. 1997, Hot ш analysis for cc
servation of plant biodiversity in the Mediten rranean Ba-
sin. Ann. y Bot. Gard. 84: 112-127.
—— . P. Linder. 1998. The phylogenetic po-
sition of Prionium (Juncaceae) within the order Juncales
based on —— ‘al and rbcL sequence data. Syst.
Bot. 23: 43-5
Oliver, E. G. н. 30 JO. Systematics of Ericeae (Ericaceae:
Ericoideae), species with indehisc - and partially de-
hiscent m Contr. Bolus Herb. 19: 1-483.
Olmstead, R. G., C. . D. Wolfe, N. D.
Nelson, W. J. Elisens & P. , „Жаа. 2001. Disinte-
E e the Scrophulariaceae. Amer. J. Bot. 88: 348—
361.
— T. D. & B. T. Styles. 1975.
graph of y bs Es ‘eae. Blumea 22:
Raven, P. H. . Axelrod. 1978. — relation-
ships of de к flora. Univ. Califumia Publ. Bot.
12
W. TS
generic mono-
540.
12.
Rebelo, А. С. 1992, Preservation of biotic diversity. Pp.
308—344 in R. M. “айан, (editor), The Ecology of Fyn-
A — Univ. Press, Cape Town.
Rutherford, C. € R. H. Westfall. 1994.
southern ve а: An objective categorization. Mem. Bot.
Africa ӨЗ.
V., M. К. Fay, D. €. Albach, A. Backlund, M.
.. M. Cameron, S. j Johnson, M. D.
Pintaud, М. Pow Sag
Soltis, Р S s ж Wes
LC e
4 neat rly a te
E gene
E.,
Biomes of
‚К. J.
2000. Phyla ny of the eu-
familial analysis based on
sequences. Kew ui 55: 257-309
Schatz, G. P. P. Lowry, M. Lescot, A. E. Wolf, S. An-
driambololonera, V. Randrinasolo &J. jos ысый
ona. 1996. Conspectus of the vascular plants of Mada-
gascar: A taxonomic and | conservation electronic data
base. Pp. 10-17 in L. +. van der Maesen et al. (ed-
itors), The Biodiversity a African Plants. Proceedings
of the XIV AETFAT Congress, Wageningen, The Neth-
erlands. Kluwer Academic Publications, Dordrecht.
| upper lacustrine
— — Banke, Namaqualand.
African pre : 1-10
Sine erie — B. R. “The — of the montane
— enic ele in the South — invertebrate
. Ann. Cape Prov. Mus. 2: 190-205.
тшш. A. 1986. Floristic Re ol as of the World. Univ.
— Press, Berkeley. [Translated by Y Crovello. |
Taylor, 1‹ 171-229 in М. J. A.
Wer “reer n Biogeography = Ко ology of Southern
. 1. W. Junk, The Hag
18. Capensis. Pp.
7
-
>
—
Afr
Tilman, jo 1082. Resource Compe * dd Community
eto
Structure. Princeton Univ. Press, Princ
1983. Some thoughts on — "competition
and — in plant communities. Pp. 322-336 in F.
J. En uger, D. E Mitchell & U. M. J: editors),
Vpe » Ecosystems, The Role c nf Nantes
анан i Berlin.
— ———, V. J. Bond, B. M. Campbell, F. J. Kruger, н. J
Linde >, A. Scholz, H. C. Taylor & M. Witter, 1983.
igin and maintenance of plant species diversity. Pp.
125-135 in J. A. Day — Mineral Nutrients in
Mediterranean Ecosystems. S. African Nat. Sci. Progr.
Rep. 71. CSIR, Pretoria
Villagran, С. М. 1994. Quarternary history of the medi-
lerranean vegetation of Chile. 3-20 in M. T. Kalin
Arroyo, P. H po & M. D. ы Гете Ecology and
Biogeography f Mediterranean Ecosystems in Chile,
таат. fornia and ain s Ecol. Stud. 108. Springer- Ver-
lag — Vork.
W ishc u, I. C., . Rosenzweig, L. Olsvig-Whittaker &
A. Se — 2000. What naka 's nutrient-poor mediter-
ranean heathlands so rich in plant diversity. Evol. Ecol.
Res, 2: 935-955.
Vogel, S. 1954. Blüte — pen als Elemente der
Sippengliede s Bot. Stud. -338.
Wagner, W. L. 1991, E — еу
son of the avai and M
floras: A compari-
Marquesan Ar
267-284 in E. C. Dudley (editor), '
lutionary Bue Vol. 1. Dioscorides Press, Portland,
айы
1990. Manual of
Hawaiʻi Press,
Herbst € S. Н. Sohmer.
the Flowering Plants of Hawaiʻi. Univ.
Honolulu.
Weimarck, Н. 1941. Phy roups, centers
ss — rar the — flora. Аш ta Univ. Lund.
ne үз 5): -M3
Xiang 0.) — Ой» D. R. Morgan & P. S. Soltis.
293. TS e N O e L. sensu lato
om lr cata relatives inferred od — sequence ale
. Missouri Bot. Gard. 80: 723—7
Volume 89, Number 2 Crosby 303
2002 Statistical Summary
STATISTICAL SUMMARY OF SOME OF THE ACTIVITIES IN THE MISSOURI BOTANICAL GARDEN HERBARIUM, 2001
Vascular Bryophyte Total
Acquisition of Specimens
Staff Collections 31.270 1.494 38,704
Purchase 201.004 35.000 236,004
Exc ang 23.088 1.812 24.900
Gif 7,837 1.705 9.542
Total acquisitions 263.199 46.011 309.210
Mountings
Newly mounted at MO 70,177 23.093 93.270
Specimens mounted when acquired 14.004. 0 14.004
Repairs
Specimens repaired 13.482 n/a 13.482
Specimens stamped 1.566 n/a 1.566
Total repairs 15.048 0 15.048
Specimens sent
On exchange 22.102 160 22.262
As gifts 19,242 028 19.870
Total 11.344 788 12.132
Loans sent
Total transactions 359 38 397
Total specimens 22,749 2,519 25.020
To U.S. institutions
‘Transactions 192 23 215
Specimens 13.028 1.948 14.976
To foreign institutions
"Transactions 167 I5 182
Specimens 9,721 931 10,652
To student investigators
Transactions 50 13 03
Specimens 6.772 1.207 7.979
To professional investigators
Transactions 309 25 334
Specimens 16.161 1.672 17.833
Loans Received
Transactions 233 28 261
Specimens 23.980 2.882 20.862
From U.S.A. From abroad Total
Visitors 234 92 326
During 2001, 108,840 — were accessioned into the herbarium: 93,270 mounted at MO, 14. 004 mounted
when acquired, and 1566 old MO
specimens in the herbarium on 1 January 20( EE 5,219,2
Purchases in 2001 included а
Missouri Bot. Сага.
repaired as needed, stamped, and accessioned in the near future.
bryophytes packeted, and all accessioned as re
Herbarium (see
specimens — — Bog red).
Bull. 89(6): hs 2001).
sources allow
The
4,832,175 vascular
. In addition, the bryophyte collection from LA
total number of mou
plants and 387,041 bryophytes).
100 vase plants and 35,000 stil in the Clyde E Reed
The estimated 162,000 mounted vascular plants should be
The unmounted vascular plants will be mounted. the
inte xd. accessloner
К, approx-
imately 30.000 specimens, accumulated and curated by William Dean Reese (1924-2002). was acquired on permanent
loan in late 2001.
304 Annals of the
Missouri Botanical Garden
The Garden’s herbarium is closely associated with its database management system, TROPICOS (see (www.mobot.org/
mobot/research)). The charts below summarize some of the statistics from TROPICOS both for the calendar year 20(
and as year-end totals. Note that the specimen records in TROPICOS are primarily based on MO specimens, meaning
that about thirty-seven percent of the bryophytes (an increase of about six percent over 2000) and thirty-two percent
of the vascular plants (an increase of about one percent) in the herbarium are now computerized, with an overall total
of about thirty-two percent (an increase of about one percent).
TROPICOS records—Calendar Year 2001 Additions
Bryophytes Vascular Plants Total
Specimens 30,709 94,277 124,986
Names 931 25,453 26,384
Synonyms 872 12,097 12,969
Distributions 369 31,376 31,745
Types 152 19,040 19,192
Bibliography 1,170 3,605 4,775
TROPICOS records—Year-End 2001 Totals
Bryophytes Vascular Plants Total
Specimens 143,298 1,548,109 1,691,407
Names 100,431 795,401 895,832
Synonyms 63,415 392,242 455,057
Distributions 39,028 798,705 837,793
Types 7,983 278,617 286,200
Bibliography 23,274 04,654 87,928
Specimens in herbarium 387,041 4.832.175 5.219.216
Percent of specimens
computerized 27 22 32
In TROPICOS, literature-based Synonymy is always linked to a reference in Bibliography and directly with at least
two records in Names, the synonym, often a basionym, and the correct name according to the reference. Additional
synonymy may be derived from these direct links, e.g., all other combinations of a basionym treated as a synonym of
a given name are also synonyms of it.
—Marshall R. Crosby
Volume 89, Number 2, pp. 125-304 of the ANNALS OF THE MISSOURI BOTANICAL GARDEN
was published on July 10, 2002.
ANNALS OF THE MISSOURI BOTANICAL: GARDEN
ATUDMISSOCUREP BOTANICAE GARDEN ANNUAL REPORT
ARE NOW AVALTLABLE UN: J STOR!
JSTOR, a not-for-profit organization, is an important endeavor
dedicated to helping the scholarly community take advantage of advances
in electronic technologies. The JSTOR Archive makes available the
complete backfiles of Annals of The Missouri Botanical Garden,
Missouri Botanical Garden Annual Report (1890-1912), and more than 100
other scholarly journals to researchers through participating libraries.
JSTOR has recently announced the release of its Ecology & Botany
Collection. Developed in cooperation with the Ecological Society of America,
the Collection includes the archives of twenty-nine journals, offering
online access to more than 1 million pages of important historic literature.
The Ecology & Botany Collection
American Journal of Botany
American Midland Naturalist
American Naturalist
Annals of the Missouri Botanical Garden
— Review of Ecology and Systematics
Conservation Biology
Diversity and Distributions
Ecological Applications
Ecological Monographs
Ecology
Evolution
Functional Ecology
Global Ecology and Biogeography Letters
Information regarding JSTOR is available at http://www.jstor.org
120 Fifth Avenue, New York, NY 10011
International Journal of Plant Sciences
Journal of Animal Ecology
Journal of Applied Ecology
Journal of Biogeography
Journal of Ecology
Journal of Tropical Ecology
Journal of The Torrey Botanical Society
Limnology and Oceanography
Missouri Botanical Garden Annual Report
Quarterly Review of Biology
Systematic Biology
Systematic Botany
viiam —
|
|
|
I
|
|
i
|
|
|
||
|
|
CONTENTS
Conservation, the 47th Annual Systematics Symposium of the Missouri Botanical Garden
Introduction: Conservation and the Future of Life
George E. Schatz & P. Mick Richardson
The Seven Great Questions of Systematic Biology: An Essential Foundation for
Conservation and the Sustainable Use of Biodiversity 2 Joel Cracraft
Taxonomy and Herbaria in Service of Plant Conservation: Lessons from Madagascar's
Endemic Families _ $ George E. Schatz
Chicago Wilderness: A New Force in Urban Conservation
Debra K. Moskovits, Carol J. Fialkowski, Gregory M. Mueller & Timothy A. Sullivan
Safeguarding Species, Languages, and Cultures in the Time of Diversity Loss: From
the Colorado Plateau to Global Hotspots
- Gary Paul Nabhan, Patrick Pynes & Tony Joe
The United States Naturalized F he — se He of Deliberate Introductions
hard N. Mack & Marianne Erneberg
The Dodo Went Extinct (and Other Ecological ph Stuart L. Pimm
The Global 200: Priority Ecoregions for Global Conservation
avid M. Olson & Eric Dinerstein
El Género bias (Acanthaceae) en Sudamérica Austral — _ Cecilia Ezcurra
Plant Diversity of the Cape Region of Southern Africa — Peter Goldblatt & John C. Manning
Statistical Summary of Some of the Activities in the Missouri Botanical Garden Herbarium,
2001 . Marshall R. Crosby
— Cover illustration, Breonia richardsonii Razafim., drawn by Barbara Alongi.
Annals
of the
Missouri
Dotanical
Garden
2002 Y
Volume 89, Number 3
Summer 2002
Annals of the
Missouri Botanical Garden
The Annals, published quarterly, contains papers, primarily in systematic botany,
contributed from the Missouri Botanical Garden, St. Louis. Papers originating out-
side the Garden will also be accepted. All manuscripts are peer-reviewed by quali-
fied, independent reviewers. Authors should write the Managing Editor for informa-
tion concerning arrangements for publishing in the ANNALS. Instructions to Authors
are printed in the back of the last issue of each volume and are also available online
at www.mobot.org/mbgpress.
Editorial Committee
Victoria C. Hollowell
Editor,
Missouri Botanical Garden
Amy Scheuler McPherson
Managing Editor,
Missouri Botanical Garden
Diana Gunter
Associate Editor,
Missouri Botanical Garden
Aida Kadunic
Senior Secretary
Barbara Mack
` Administrative Assistant
Ihsan A. Al-Shehbaz
Missouri Botanical Garden
Gerrit Davidse
Missouri Botanical Garden
Roy E. Gereau
Missouri Botanical Garden
Peter Goldblatt
Missouri Botanical Garden
Gordon McPherson
Missouri Botanical Garden
P. Mick Richardson
Missouri Botanical Garden
Charlotte Taylor
Missouri Botanical Garden
Henk van der Werff
Missouri Botanical Garden
For subscription information contact ANNALS
OF THE MISSOURI GARDEN, % Allen Marketing
& Management, P.O. Box 1897, Lawrence, KS
66044-8897. Subscription price for 2002 is
$140 per volume U.S., $150 Canada & Mexico,
$175 all other countries. Four issues per vol-
ume. The journal Novon is included in the sub-
scription price of the ANNALS, —
annals@mobot.org (editorial queries)
http://www.mobot.org/mbgpress
O Missouri Botanical Garden 2002
THE ANNALS OF THE MISSOURI BOTANICAL
GARDEN (ISSN 0026-6493) is published quar-
terly by the Missouri Botanical Garden, 2345
Tower Grove Avenue, St. Louis, MO 63110.
Periodicals postage paid at St. Louis, MO and
additional mailing offices. PosrMASTER: Send
address changes to ANNALS OF THE MISSOURI
BOTANICAL GARDEN, % Allen Marketing &
Management, P.O. Box 1897, Lawrence, KS
66044-8897.
The mission of the Missouri Botanical Garden is to discover and share knowledge about plants and
their environment, in order to preserve and enrich life.
€ AKICI/ATIC[(. 740 40 IRAN РТУ.
: `
© Thi:
ең
at Dríinmor)
Ls /
olume 89 Annals
umber 3 of the
002 Missouri
Botanical
Garden
PHYLOGENETIC Ana María Cialdella? and
RELATIONSHIPS OF THE Liliana Mónica Giussani?
GENUS PIPTOCHAETIUM
(POACEAE, POOIDEAE,
STIPEAE): EVIDENCE FROM
MORPHOLOGICAL DATA'
ABSTRACT
A systematic treatment of the entire genus Piptoc — — h complements ce of Cialdella and Arriaga for the
South American spec les, 18 presente d * теш. A synopsis of : spec ies and 2 varieties, a key to ident tify these 5| ECTES,
as well as synonymy and geographic ‘al distribution, are given. Hlustrations and E тсн of di gnostic characters
are presented for the North American species = the first time. We propose a phylogenetic hypothesis for the genus
based on morphological and anatomical characters. Monophyly " Piptochaetium was supported by the combination of
three synapomorphies: a grooved palea. aaa lemma margins that fit into a ол Шла palea groove, and the palea
protruding from the lemma. Two sections were previously recognized in Piptochaetium: Podopogon and Piptoc haetium,
The cladistic analysis showed that only South (American species of Pipt vi opc sect. Piptochaetium, together with /
brevicalyx, form a monophyle lic group (Obovoid а while Piptochaetium sect. Podopogon is polyphyletic. Therefore,
the taxonomic recognition of sectional taxa cannot be supported. Phylogenetic relationships of two varieties of Pipto-
chaetium stipoides (var. stipoides and var. echinulatum) were nol Kn resolved. Furthermore, based on cladistic
analyses, taxonomic observations, and the study of type material. Piptoc haetium tovarii subsp. pilosa is raised to specific
rank (= P. pilosum).
Key words: phylogenetic relationships, Piptochaetium, Poaceae, Stipeae.
Rest MEN
* presenta un tratamiento sistemático de todo el género Piptochaetium, que e omple menta el propuesto por Cialdella
dia para las especies sudamericanas. En este — se presenta una sinopsis de las 36 e * ‘cies y 2 variedades,
una с have para identificar las especies, su sinonimia. distribución geográfica y observaciones. Se incluy or primera
vez ilustraciones у fotograffas de los caracteres di agnóstic os para las « "spec les norteamericanas. Se propone una hipóte SIS
' This research was supported by CONICET, PEL: 0000/97. We thank O. Morrone, F. O. Zuloaga, M. Barkworth, J.
Maze, and two — reviewers who kindly reviewed versions of the manuscript and guided us with hair construc-
tive comments, and G. Davidse and R. Soreng who helped us to solve nomenclatural queries. We are also grateful to
the curators and staff di the herbaria mentione di in the text, who allowed examination of specimens in their care, whether
through loans or during personal visits. Thanks to Daniel Rodriguez for assistance in the use of the scanning microscope,
and Vladimir ro Dudás for preparing the line drawings and the composition of the photographic — F Sem we thank
the staff » the Darwinion, who helped us — *—
? Instituto de Botánica calera m Labardén 200 CC 22 (B1642HYD) San Isidro. Buenos Aires. Argentina.
acialdella@darwin.educar
ANN. MISSOURI Bor. GARD. 89: 305-330. 2002.
306
Annals of the
Missouri Botanical Garden
filogenética basada en
nofilia de Piptochaetium sobre la base de tres sinapomorfías: una pea surcada entre los
lemma involutos insertos en el surco longitudinal de la pálea y la pálea
caracteres morfológicos y anatómicos. Mediante el análisis filogenético, fue confirmada la mo-
nervios, los márgenes de la
asomando por entre los márgenes de la lemma.
Dos secciones fueron previamente reconocidas para Piptoc haetium: Кол. у — haetium. Como resultado del
análisis cladfstico,
*
miento de categorfas formales infragené ricas. En e
sti ae y var. ec Шш im) у la posición а де aw variedades no fue clara
e (= P. pile
p
tovarii subsp. pilosa fue elevada al rango de espec
vaciones taxonómicas y el estudio del — *
ólo las especies sudamericanas de la sección
un grupo — © (grupo Obovoide). Consecuentemente, en este tratamiento, no se quani ni vá
1álisis, se incluyeron 2 variedades de Piptochaetium stipoides (va
‘on P brevicalyx, conformaron
álido el reconoci-
Piptochaetium, junto
a. Piptoc hasten
жит), sobre la base del estudio filogenético, las obser-
mente resuelta
Piptochaetium J. Presl is an American genus dis-
tributed mainly in temperate areas of both hemi-
spheres and scarcely represented in tropical re-
gions of the Andes (Parodi, 1944; Barkworth, 1986;
Sánchez Vega, 1991; Cialdella & Arriaga, 1998).
The genus belongs to the tribe Stipeae and
established by Presl in 1830, based on Piptochae-
tium setifolium J. Presl |= P. panicoides (Lam.) E.
Jesv.|. It includes perennial, caespitose plants,
generally less than 1.5 m . with loosely or
densely flowered panicles of one-flowered spikelets:
membranous glumes, frequently longer than the flo-
ret, enclose the anthecium. Three characters are
useful for recognizing the genus: the grooved palea,
involute lemma margins that fit into a longitudinal
palea groove, and the palea protruding from the
lemma.
The South American species of Piptochaetium
have been studied by Parodi (1944), who treated
21 species in 2 sections: Piptochaetium sect. Po-
dopogon (Raf.) Parodi and Piptochaetium sect. Pip-
tochaetium (= Piptochaetium sect. Eupiptochaetium
Parodi). Sanchez Vega (1991) treated the 6 species
for Peru, and grouped them into the sections pro-
posed by Parodi. Cialdella and Arriaga (1998) pub-
lished a more recent treatment of the South Amer-
ican species, in which they recognized 27 species,
2 subspecies, and 2 varieties for both sections, con-
sidering leaf anatomy and micromorphological
characters of the lemma in their analysis. T
North American taxa were studied by Valencia and
Costas (1968) with emphasis on eytological char-
A compari-
acters, and they recognized 7 species.
son of these treatments, including the infrageneric
is shown in Table 1. In a revision of
Barkworth (1986)
and
classification,
the North American. Stipeae,
North
without grouping them into sections.
Mesoamericas
=]
treated 13 species ir
The aim of this work is to complete the morpho-
logical account of the genus by considering the spe-
cies not included in the previous paper by Cialdella
and Arriaga (1998).
opsis of 36 species and 2 varieties, a key to identify
This study comprises a syn-
them, and their associated synonymy and geograph-
ical distribution. A phylogenetic analysis of Pipto-
chaetium was performed to explore the relationship
among species, using cladistic methodologies with
the morphological data.
MATERIALS AND METHODS
MORPHOLOGICAL STUDY
This study was based on the analysis of almost
1000 herbarium specimens, including type material
for most of the analyzed taxa (Appendixes 1, 2).
The specimens studied were from BA, BAA, BAB,
BAL, C, CAS, CEN, CORD, CPUN, CTES, G, ICN,
K, LE, LP, LPB, MO, MEXU, MVM. MVFA, NY,
P, PR, SGO, SI, SRFA, UNR, US, USM, UTC and
XAL (Holmgren et al., 1990).
Micromorphological structures of the lemma
were analyzed using a scanning microscope Zeiss
DSM 940 A, located in the — de Botánica
Darwinion and as summarized in Cialdella *
Arriaga (1998: 108). Epidermal ndis (ma-
crohairs, prickles, hooks, and papillae) follow the
classification proposed by Ellis (1979).
CLADISTIC ANALYSIS
Twenty-nine exomorphological and epidermal
characters were used in the cladistic analysis (Ap-
pendix 3); all multistate characters were treated as
unordered. Only two autapomorphies were included
in the analyses: characters 20 and 27 (Cialdella &
Arriaga, 1998: figs. 3D and 9F, respectively). Char-
acters were directly observed from herbarium ma-
terial totaling nearly 1000 specimens. The data ma-
trix is presented in Table 2.
Two different computational strategies, using min-
imum parsimony, were used to analyze phylogenetic
relationships among species of Piptochaetium. In
both cases, polymorphic characters were included
Table 2). One strategy assigned all characters equal
3
weight. This analysis was performed using NONA
version 1.6 (Goloboff, 1993a). In the second strategy,
a concavity function of the homoplasy (K = 3) was
Volume 89, Number 3 Cialdella & Giussani
2002 Phylogenetic Relationships of Piptochaetium
[|
Figure 1. SEM photos showing epidermal charac ter rs of the lemma of Piptochaetium.—A,. Lemma without any epider-
mal — s of P. brevicalyx (Rzedowski 35863 ‚ Macrohairs of P. fimbriatum (Davidse et al. 29775). —C. Prickles
and hooks of P. avenaceum (Hart ey 1654). —D. Pric kle s, hooks, and macrohairs of P angustifolium (Palmer 720).
y
used to apply weight to all characters, a method for nus from its allies (Spegazzini, as Oryzopsis Michx.
estimating implied weight (Goloboff, 1993b). This sensu Speg.. 1901: Parodi, 1944: Barkworth, 1990:
analysis required Pee-Wee version 2.8 (Goloboff. Peñailillo, 1996; Cialdella & Arriaga. 1998). Other
1997). In both analyses, settings amb- (clades re- characters were also used to uniquely identify Pip-
solved only if they have unambiguous support). and tochaetium, such as a palea longer than the lemma.
poly = (polvtomies allowed) were used to resolve protruding from the lemma apex, and the lemma
phylogenies. The order of taxa was randomized. cre- margins involute and fitted into the palea. groove.
ating a Wagner tree and submitting it to branch- The 37 entities recognized within the whole genus
swapping by means of tree-bisection reconnection were considered as terminal taxa.
(TBR) using the sswap* command. To evaluate the
relative support of clades, branch support. with in- OUTGROUPS
struction "bsupport" (bs) of NONA (Bremer. 1988.
1994), was calculated.
Vassella E. Desv. and Hesperostipa (Elias) Bark-
worth were selected as outgroups, based on their
— similarities in morphological and anatomical char-
acters to Piptochaetium (Thomasson, 1978b: Bark-
Since Desvaux (1853), Piptochaetium has been worth, 1990) as well as previous phylogenetic re-
recognized by the presence of a grooved palea, and lationships proposed by Barkworth (1990).
this character was diagnostic to distinguish the ge- A detailed history of Nassella can be found
Annals of the
308
Missouri Botanical Garden
11 unmqní y wnyognl q wnyognl 4
чо av miri d итп d итү d
sn AN n9 wnyouquy y UNIDUQUAS y
11 штр]үпәпә Y ungpjmono y
1n uy SUdISAYDI q SUdISAYDI Y SUIIAYDI J
тү WNUDIADYING Y WNUDIADYING + WNUDILDYANG q
aN x4qD21424Q. 41 XA]DINIALG f xona qd
yu) әѕиәјодир 4 asuajozun + ISUIJOTUD J
aN umpofusnduo y
dnoa:) pro^oq() WALL V HOOLdld I ALEAV HOO. dld WALLAV HOOLdld I OLET V HOOTdld
9A “OW “ney SUAISALVA d SUIISALTA d SUIISIL1A d
aN ‘Nd мә]ә$ ү
ag mndojspavs J nnZojsp3ps y INIASDIDS 4
1Q ug чү wnupiyyooidns q unupj224dn4. y unupnp[224dna J
1 N 1ә#ина q
19 alsniod q ənsnpd y
ү asuapnjsodpu y asuapjsodpu J ISUIDISOÍDU j
1Q ag чү штрәш + штрәш y 10102 4 =
3g 53 09 чү unmpu 4 штпри q unmpu 4 шптри ¢
үчү е а 3 Шет А PY qq
sa AN по) uunjpaquiaf 4
әд 19U01819YIDOS J 121018124102 ү] 1әио]кәуттә 41
1 uy wnsnfuod q wnsnfuor x] umsnfuoo y
IV 2D4240D2 J 2D421QD2 d 2012192 4
ү vNULIASAYIDAQ 4 umuladsíyopaq Y штица f
11 Чочу Dd орол M озн О
sq ѕәр1отәриәпр d sapio1oDuaap d P sapioi0DuodpD d
Sn N 1unaopuaan qq UNIIDUIAD + unaonuoap yJ
aN umnofisnzup q 2 ungofiusnzup y
19 unuidp q und. q
пол ons[sudsqoq NODOdOGOd NODOdO(GOd NODOdOGOd NOODOdOGOd
uonnquisip sisA[eue orouozo| sud (8661) (1661) (8961) (rr61) !poreq
jeorydeis0a+) əy} woj} e3eLLIV Y еәрүегә
sdnois [ешлоуи
PBA zaqoueq
581507) DY EIOUS[EA
nsuas suog
'pajuasaud OS[B SI 591 ээ әш jo uornqugjstrp orgde1302^) ‘SISA[BUE 294330] &ud 34] шол 51891 Чим uosueduio» ut umapyooidid ишцим SUOTJBOYISSP]) опәпәЗр1уи]
“B[PNZIUDA “OA S81 Pau]
“Sn ‘Aenanı 1] "un 1d “Əd ‘Aene qf “ed :OJIXIJA “OW `РТРШӘ1РП С) пс) чорепоу EX 1214шо[07) *07) U) чэ: 11219 ug :BIAIOH "og teurnuaz1y AY :0] puodsaunioo ѕиоцет, A2Jqq y
т 1L
Volume 89, Number 3 Cialdella & Giussani 3
2002 Phylogenetic Relationships of Piptochaetium
Continued.
Table 1.
a 3 | Barkworth (1990). She expanded Nassella, includ-
= ec C Р - ing a total of 79 species, the majority of which were
= ESS = б previously considered in Stipa L. The cireumscrip-
E — — p 5 ye 5 " 2 tion proposed by Barkworth (1990) was later ac-
= Sx gua pS 4 со CP а iu cepted by Peñailillo (1996) and Torres (1997). In
gost ee به E SHP ې ы this sense, Nassella includes species with obovoid
| oia or terete florets, a short and blunt to long and sharp
3 : callus, strongly overlapping lemma margins, a lem-
, A = + ma apex fused in a crown, and a highly reduced,
= E — nerveless and glabrous palea. All species consid-
be 3 y е ie ered in Nassella have a characteristic lemma epi-
as t Eg 5 © 2 2 » dermal pattern: short fundamental cells, with irreg-
57 Ет 8 = £ Š i "E ularly sinuate sidewalls and abundant short cells.
= E * = E E B. Ё & & E E Hesperostipa was one of the nine sections of Stipa
ш р E Bas * bie = A = -— formally proposed by Elias (1942). Thomasson
hdc co M M, (19784) noticed a very different epidermal pattern
within this section in comparison with any other
E Stipa: long fundamental cells with strongly sinuous
3 Fe E sidewalls, and absence of short cells. After studying
= Е $ E North American Stipeae, Barkworth (1993) raised
2. = T2 а Stipa sect. Hesperostipa to the generic. rank, in
T 2a H8 . E Su 4 which she included four species: Hesperostipa co-
28 232 i в Е ti р : mata (Trin. & Rupr.) Barkworth. H. curtiseta
Bc E à چ E E E 3 A (Hitehe.) Barkworth, H. neomexicana (Thurber)
= ¿72 S x ст а § Barkworth, and H. spartea (Trin.) Barkworth. Hes-
E паа. E oO А АРЕ ll a; perostipa 1s endemic to North America and includes
plants with narrowly terete florets, a sharp callus,
and a persistent, twice-geniculate awn. The lemma
m is indurate, with flat margins, slightly or not over-
3 S lapping: the palea is subequal to the lemma, prow
& z lipped and 2-nerved (Barkworth, 1993).
p z Š i z е KEY TO PIPTOCHAETIUM, NASSELLA, AND HESPEROSTIPA
5 = х © = = Jl Palea longitudinally grooved, 2-nerved, longer
B = = a x than the lemma, and SEU from the lemma
x x \ \ apex; lemma margins involute, fitting into the
palea groove Piptochaetium
1'. Palea flat, 2-nerved or nerveless, shorter than
or equal to the lemma, usually not protruding
beyond the lemma apex; lemma margins fl: 2
Lemma margins strongly overlapping; pa-
n lea nerveless, less than one-third the
E length of the lemma lassella
em а 2'. Lemma margins not or slightly sonia»:
ge E E ping: palea. 2-nerved, always as long as
E в) = the lemma, apex prow tipped or
© * “pine A Hesperostipa
Е: xX
5
E RESULTS
3 E Seven equally parsimonious trees were obtained
— = = when the characters were weighted equally. These
z= - 8 : Ы were 58 steps long. with a С — y Index (CI) =
2 = E 5 = i ёв : < 0.55 and Retention Index (RI) = = 0.87. When the
| چ s 3 ES ГРЕЕ So analysis was performed without autapomorphies
F & ÉSE = E X E 3 E (characters 20 and 27), the CI was 0.53. When the
eee A ECA x characters were weighted, five equally parsimonious
310
Annals of the
Missouri Botanical Garden
Table 2
Data matrix used in the cladistic analyses for Piptochaetium. Numbers in the first row represent the
characters described in the text, together with their codification, Numbers within brackets represent a polymorphism
for the particular character of the species involved.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Nassella 0 0 2 ооо [OI] [01] [01] [O I] [01] [023] [01] [01]
Hesperostipa 0 0 0 1] 1 2 l 0 1 0 [0 1] 0 [0 1]
Piptochaetium alpinum l d 1 1 l 1 l 0 1 0 1 0 0 l
P. angolense 1 1 1 1 1 1 0 | 1 1 0 0 2 1 0
P. angustifolium Il dd d 1 1 1 0 0 ] 0 0 0 0 1
P. avenaceum |I 1 | 1 .) !] 1 0 1 0 JO 1] 0 0 0 0
P. avenacioides | 1 d 1 1 !] l 0 1 0 0 0 0 0
P. bicolor 1 1 | 1 | ! 1 0 1 0 1 0 1 0 0
P. brachyspermum Г l l 1 l 1 1 0 1 0 l 0 l 0 0
P. brevicalyx Г 1 1111 0 l 0 1 ) 1 2 0 0
P. burkartianum I 1 1111 0 l 1 l [01] 0 2 0 0
P. cabrera 1| 1 l 1 | 1 l 0 l 0 0 0 0 0
P. calvescens |I] 1 1 J ! 0 l 0 l 0 0 2 0 0
P. confusum 1 1 | 1 1 ! 1 0 1 0 1 0 0 0 0
P. cucullatum l 1 |1 1 |1 1 0 l 0 l 0 0 2 0 0
P. featherstonei l 1 | 1 1 ! 1 0 1 0 0 0 0 0 1
P. fimbriatum | 1 | 1 |1 1 0 l ] 0 1 0 0 0 1
P. hackelii |o 1 | 1 |1 1 l 0 1 0 l 0 0 0 0
P. hirtum 1 1 d 1 l | 0 І 0 | 0 0 2 0 0
P. indutum 1 1 l l 1 1 l 0 1 0 0 0 0 0 1
P. jubatum 1 d 1 og) l 1 0 | 1 | JO 1] 0 2 0 0
P. lasianthum l l 1 1 1 | 0 | l l l 2 0 1
P. lejopodum 1 1 | |1 1 1 0 l 0 | 0 0 2 0 0
P. medium | 1 | l 1 1 l 0 1 0 l l 0 0
P. montevidense | Ж 1 ]) ].» I 0 1 0 l [01] 0 3 1 0
P. napostaense | 1 |1 1 11 | 0 l 0 0 0 0 0
P. palustre | 1 1 1I 1 !]| | 0 1 0 1 0 0 0 0
P. panicoides | dd 1 |) l | 0 1 0 l 0 0 3 1 0
P. pilosum | i 1 1 1 d 0 | l l 0 0 2 0 1
P. pringlei 1 d gd | l | l 0 l 0 l 0 0 0 1
P. ruprechtianum 1 dd 1 |1 l | l 0 l 0 l 0 l 0 0
P. sagasteguii 1 l | 1 1l I l 0 1 0 1 0 0 0 1
P. seleri l l l gd l | 0 1 l 0 ] 0 0 0 1
P. setosum 1 1 1 | 1 | 0 1 1 1 0 2 0 0
P. stipoides var. echinulatum | | 1 1 | | 0 l l l [01] 0 2 0 0
P. stipoides var. stipoides Lor 1 |1 1 !] 0 l l l [01] 0 2 0 0
P. tovarii Lot d | 1 l 0 1 0 1 0 2 0 0
P. uruguense I t 1 | l l 0 1 l l [01] 0 2 0 0
P. virescens l i 1 1 1] 1 0 l 0 0 0 1
trees were obtained. These trees were also 58 steps
their fit was 190.7, and the
Although the trees obtained with the two strat-
long, rescaled fit was
0.68.
egies differed in some respects, both yielded the
same strict consensus tree as well as the unambig-
uous
Only
volved P. calvescens, P. cucullatum, P. jubatum, P.
synapomorphies that support clades (Fig. 3).
the relationships within one clade, which in-
lejopodum, P. stipoides var. stipoides, and P. stipoides
var. echinulatum, differed between the strategies.
Relationships among these species are not resolved
on the consensus tree (Fig. 3). Figure 4 shows one
of the seven hypotheses found with NONA. This tree
is the most similar to one of the five trees found with
Pee-Wee.
which,
They differ in the position of P. alpinum
in Pee-Wee's results, is sister to the clade
that includes P. pringlei to P. palustre; and they also
differ in the order of P. uruguense and P. angolense,
which are inverted in Pee-Wee analysis.
Cladistic analysis supports Piptochaetium as a
monophyletic genus [Bremer support (bs) = 3]. It is
clearly defined by three synapomorphies (Fig. 3):
lemma margins involute (1), a boat-shaped palea (2),
2A, B).
and a palea longer than the lemma (3; Fig.
Volume 89, Number 3
2002
Cialdella & Giussani 311
Phylogenetic Relationships of Piptochaetium
Table 2. | Extended.
l6 17 18 19 20 2] 22 5:23. 24 25 20 27 28 29
Vassella [0 1] [0 1] [0 1] 0 O [01] [01] O [01] 0 [01] O [01] 0
Hesperostipa 0 | l 0 0 0 1 0 1 0 1 0 0 0
Piptochaetium alpinum 0 l | 0 0 | l 0 1 0 1 0 0 0
P. angolense 0 1 l 0 0 l 1 0 0 0 1 0 | 0
P. angustifolium 0 | | 0 | l 0 l 0 l 0 0 0
P. avenaceum 0 1 [O 1] [01] 0 | l 0 1 0 l 0 0 0
P. avenacioides 0 1 [0.1] [OT] 0 | l 0 1 0 l 0 0 0
? bicolor 0 1 0 l 0 l l 0 l 0 1 () 0 0
P. brachyspermum 0 1 0 1 0 | l l О 0 1 0 0 0
P. brevicalyx 0 0 0 0 0 | | 0 0 0 0 0 | 0
P. burkartianum 0 0 0 0 0 | | 0 0 0 1 0 | 0
P. cabrerae 0 l | 0 0 | l | 0 0 1 0 0 0
P. calvescens [0 1] | | 0 0 | 0 | 0 l l 0 | |
P. confusum 0 l 0 | 0 | | 0 1 0 | 0 0 0
P. cucullatum 0 1 0 l 0 | l | 0 0 1 1 | 0
P. featherstonei 0 1 | 0 0 | | 0 1 0 1 0 0 0
P. fimbriatum 0 0 0 0 0 | | 0 [01] O [OT] 0 | 0
P. hackelii 0 1 0 | 0 0 | ( 1 0 1 ) О 0
P. hirtum 0 0 0 0 0 | | 0 0 0 | 0 | 0
P. indutum 0 1 | 0 0 0 | 0 0 0 | 0 0 0
P. jubatum 0 l 0 | 0 | ОТ | 0 | 0 І |
P. lasianthum 0 1 l 0 0 | 0 l 0 l 0 | 0
P. lejopodum 0 1 0 | 0 | 0 | 0 0 1 0 | 1
P. medium 0 1 0 | 0) | | 0 0 0 1 0 | 0
P. montevidense l 0 0 0 0 | | 0 0 l 1 0 | 0
P. napostaense 0 1 0 l 0 0 | 0 1 0 1 0 0 0
P. palustre ) 1 0 | 0 | | 0 l 0 l 0 0 0
P. panicoides [Ol] 0 0 0 0 | | 0 0 l 1 0 | 0
P. pilosum 0 l l 0 0 0 | 0 0 0 1 0 | 0
[ pri slei 0 0 0 0 0 l l 0 1 0 1 0 0 0
P. ruprechtianum 0 | 0 | 0 | | 0 l 0 l 0 0 0
P. sagasteguii 0 l l 0 0 0 | 0 1 0 1 0 0 0
? seleri 0 1 1 0 0 | | 0 l 0 1 0 [0 1| ()
P. setosum 0 1 0 | 0 | | 0 0 0 1 ) l 0
P. stipoides var. echinulatum 0 | | 0 | l 0 І 0 l l 0 | |
P — var. stipoides 0 [0 1] [O 1] [OI] 0 | [Ol] 1 [01] [01] [Ol] 0 | 101]
Р. tova О 0 0 0 0 0 l ( 0 0 0 0 l 0
P. uruguense 0 l 1 0 0 l l 0 0 0 1 0 | 0
P. virescens 0 1 l 0 0 0 | 0 1 0 l 0 0 Q
Hesperostipa is the most closely related group to Pip-
tochaetium, as they share a 2-nerved palea (4) and
long fundamental cells of the lemma epidermis (5).
South American species of Piptochaetium sect.
Piptochaetium, together with P. brevicalyx from Mex-
2),
which is now recognized herein as the Obovoid
ico, are gathered in a monophyletic clade (bs
eroup (Fig. 3). It is supported by two synapomor-
phies: a deciduous awn (10) and obovoid florets (13,
Fig. 5H). Three species are closely related to the
Obovoid clade: P. fimbriatum (Fig. 5G). P. seleri (Fig.
5F). and P. angustifolium (Fig. 5E). Piptochaetium
—
fimbriatum and P. seleri present a blunt or truncate
o = 1) and, together with P angustifo-
bs = 1). These
three species are not included in the Obovoid clade.
callus (8
lium, they all have a short callus (7:
as they do not have a deciduous awn as well as a
North
can species; P fimbriatum and P. seleri were also
well-defined obovoid floret. They are Ameri-
found in Guatemala, with intermediate forms ob-
served between the Obovoid clade and the rest of
the species (Figs. 3 and 4: see discussion). Species
of the Obovoid clade present the relation between
length and width of the floret always less than 3.5
312
Annals of the
Missouri Botanical Garden
Figure
(Rzedow = 35863). —B. P. fimbriatum (Davidse et al. 2
(Palmer 726
(28). However, this character is ambiguous for the
clade that gathers the Obovoid group with P. seleri
and P fimbriatum, and consequently, this synapo-
morphy is not shown in Figure 3. An additional anal-
ysis was performed to test the influence of the in-
termediate species (P. fimbriatum, P. angustifolium,
and P. seleri) on the results. Without these three spe-
cies, the major clades and relationships among the
species did not change.
All species of the Obovoid clade, except P la-
sianthum, have densely flowered inflorescences (11),
although for some species of this clade this character
is variable. These species also have a crown without
macrohairs (24), which reverts in P jubatum and is
stipoides. Species
polymorphic P. stipoides var.
from P. uruguense to P. stipoides var. echinulatum
form a clade defined by a glabrous lemma (15, Fig.
LA, 5H; bs = 1), and P. pilosum (= P. tovarii var.
SEM photos showing distal portion of * — with different crowns in Piptoc — —A.
5). —C. P. avenaceum (Harvey 185
P. brevicalyx
—p. P a
pilosa Sánchez Vega) is consequently apart from its
counterpart P. tovarú (= P. tovarú var. tovaril), sug-
gesting this rank is not appropriate. Species with a
few prickles on the lemma (18) are gathered in a
clade that includes species from P. burkartianum to
Р. stipoides var. echinulatum (bs = 1). This grouping
is divided into two major clades. The first clade, Р
burkartianum, P. montevidense, P. panicoides, P. hir-
tum, P. brevicalyx, and P. tovarii (bs =
acterized by a lemma without prickles (17, Fig. 1A),
and, except for P. burkartianum, the other species
also have a glabrous callus (9; bs = 1). Piptochae-
). is char-
tium brevicalyx and P. tovarii are related by the ab-
sence of prickles in the crown (26; bs = 1), while
P. montevidense and P. panicoides are gathered in a
clade (bs = 3) and share three synapomorphies:
lens-shaped (13) and laterally compressed florets
(14), and presence of papillae in the crown (25).
Volume 89, Number 3 Cialdella & Giussani 313
2002 Phylogenetic Relationships of Piptochaetium
Nassella
Hesperostipa
Piptochaetium sagasteguii
A
сл
E
2 P. virescens
*
7 1 P. alpinum
1
11 1 P. pringlei
0
hackelii
N
w
=
napostaense
bicolor
о = о
ruprechtianum
15 19
brachyspermum
medium
confusum
avenaceum
avenacioides
palustre
featherstonei
cabrerae
indutum
angustifolium
fimbriatum
seleri
lasianthum \
pilosum
uruguense
angolense
E
OD DAD DOT DDO TD DOT DOV OD TO YG ODO ®
burkartianum
hirtum
brevicalyx
a)
„ tovarii
O-O<000O0
P. montevidense
y
N
v
.2 €
©
© .
a
=
3
TCO Q0
24 stipoides
1 HT P. jubatum
1
20 P. calvescens
—{— P. stipoides var. )
1
echinulatum
Figure 3. Strict consensus of seven equally parsimonious trees resulting from phylogene tic analysis of "e haetium
(NONA program). A similar tree resulted from the strict consensus of five equally parsimonious trees under implied
weight (Pee-wee program). Optimization of characters are shown on the tree: solid bars — unambiguous abel vn S,
empty bars = homoplastic character transformations, solid crosses = reversals. Character numbers (see Appendix 3)
are above symbols, and character state changes are below q or Crosses. Tae сав show evolution of floret shape
in Piptochaetium (see discussion).
314
Annals of the
Missouri Botanical Garden
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIS, Nassella
LPAI NTE PE PENDAN
Hesperostipa
Piptochaetium sagasteguii
virescens
featherstonei
cabrerae
indutum
pringlei
hackelii
napostaense
ruprechtianum
bicolor
vuvuwy yD озо DHA NV
brachyspermum
. medium
. avenacioides
. avenaceum
. confusum
P. palustre
d
alpinum
angustifolium
seleri
References
#444. Species found all over North and
South America
Species with — M in
North and South Am
— found in — or
n NA and Guatemala
Nabe A
Species found in South America
Figure 4.
One of the seven equally parsimonious trees of 58
fimbriatum
lasianthum
montevidense
—
دن
panicoides
hirtum
brevicalyx
bh
эээ о о э э озозҗоҗз зл ул»
tovarii
setosum
cucullatum
jubatum
lejopodum
stipoides var. stipoides
calvescens
stipoides var. echinulatum
steps resulting from the phylogenetic analysis of
Piptochaetium (NONA program), based on morphological and epidermal characters. Distribution of species is mapped
on the tree as shown in the references. Numbers below branches represent the Bremer support value for each clade.
ateral compression of the floret (14) is a derived
character that occurs also in P. angolense (Cialdella
& Arriaga, 1998: fig. 9C). The second major clade
comprises P. setosum, P. calvescens, P. cucullatum, P.
jubatum, P. lejopodum, and two varieties of P. sti-
poides (bs = 1). This clade is characterized by the
presence of prickles in only one-third of the lemma
(19), except P. stipoides var. echinulatum, which has
prickles on almost all the lemma surface. With the
exception of P setosum, all other species have a rev-
Volume 89, Number 3
2002
Cialdella & Giussani 315
Phylogenetic Relationships of Piptochaetium
olute crown toward the outside (23; bs = 1): and
19, 22, 25, and 29
is ambiguous on this clade. а among
optimization of characters 9, 18,
species are not resolved within this clade. Only /
brachyspermum and P. cabrerae also have a —
crown, a character that appears independently
these taxa.
Species traditionally included in Piptochaetium
sect. Podopogon (Cialdella & Arriaga, 1998) result-
ed in a polyphyletic group. Species of this section
are linked to some North American species, not in-
cluded in any previous infrageneric classification
(Table
gathered
1, Fig. 3). Some species of this section are
in stable clades based on unambiguous
synapomorphies: Р, featherstonei, P. cabrerae, and P.
indutum have loosely flowered inflorescence branch-
1; bs = 1). while P. cabrerae and P. indutum
bs = 1).
Piptochaetium cabrerae is characterized by a gla-
brous lemma (15) and a revolute crown (23). A clade
including P. pringlei (Fig. 5D) to P. palustre is dis-
tinguished by the absence of prickles over almost
the entire lemma surface (18: bs = 1).
es (1
do not have macrohairs in the crown (24:
All species
of this clade (bs = 1), except P. pringlei, have a
glabrous lemma (15). with prickles restricted to the
distal third (19). Piptochaetium hackelii and P. na-
postaense are identified by a lemma as wide as the
bs = 1). This character appears as re-
versals in P. indutum, P. pilosum, and P. tovarit. Four
crown (21:
species, P. bicolor, P. brachyspermum, P. medium.
and P. ruprechtianum, are gathered in a clade (bs =
1) supported by the shape of the floret: all of them
have obconical florets (13), rather than a terete to
fusiform floret. Piptochaetium brachyspermum and P.
medium are distinguished by the absence of macro-
hairs in the crown (24; bs = 1). Only Р medium has
a wide floret with respect to its length (28. Cialdella
& Arriaga, 1998: fig. 8B).
Only P. cucullatum (Cialdella & Arriaga, 1998:
fig. OF) presents a cone-shaped awn (27). and P
stipoides var. echinulatum (Cialdella € Arriaga,
1998: figs. 3D, 10J) has prickles with their bases
covered by adjacent epidermal cells (20). Pipto-
chaetium brevicalyx and P. alpinum present the up-
per glume as long as the lemma (12). although this
character has an independent origin in both taxa.
DISCUSSION
Piptochaetium is a monophyletic genus clearly
defined by reproductive characters related with
lemma and palea structures (Fig. 3). The origin of
Piptochaetium was supposed to be in North Amer-
ica according to Thomasson (1978a. 1980). who re-
lated this genus with Berriochloa Elias as its an-
cestor. This fossil grass was first characterized by
Elias in 1932. 1985) found
some similarities between Piptochaetium and Be-
Thomasson (1978a, b.
rriochloa based on floret shape. which can be cy-
lindrical, obovoid, obconical, or lens-shaped in
both genera. Although he emphasized. similarities
between epidermal patterns of the lemma, Berrioch-
loa, Piptochaetium, and Hesperostipa (treated as Sti-
pa sect. Hesperostipa by Thomasson) share a similar
lemma epidermis with long fundamental cells with
1978a, b. 1982).
This study shows Hesperostipa to be more closely
related to Piptochaetium than Nassella. A
comprehensive evaluation of
sinuous sidewalls (Thomasson,
more
ص
the relationship «
Piptochaetium to other genera in the tribe requires
a more inclusive study.
Within Piptochaetium, species were grouped in
two sections: Podopogon and Piptochaetium (Table
1) according to the revision carried oul by Parodi
(1944) and as later accepted by Valencia and Cos-
tas (1968). Sánchez Vega (1991). and Cialdella and
Arriaga (1998
As suggested by cladistic analysis, Piptochae-
tium sect. Podopogon is not a natural entity. Con-
sequently, infrageneric ranks cannot be supported.
The species traditionally included in Piptochaetium
sect. Piptochaetium (Parodi, 1944). with the excep-
tion of P. fimbriatum and P. pringlei. were assem-
bled
called the Obovoid group. also including P. brevi-
—
here
1
-
a monophyletic group. informally
calyx. This clade is supported by two synapomorp-
The Ob-
ovoid clade shares other synapomorphies with three
hies: a deciduous awn and obovoid florets.
North American species: a short callus with P. an-
gustifolium and a short and blunt callus with P.
fimbriatum and P. seleri. Short and blunt calluses
are derived states within the genus, while acute and
Fig. 3). This
who found that all
the floret
shape. always had an acute callus. He suggested
long calluses are plesiomorphic states
—
agrees with Thomasson (1985),
species of Berriochloa, regardless of
that the evolution of the blunt callus in Piptochae-
tium is a post-Miocene event.
Transition in floret shape was observed by Tho-
1978b) Ne-
braska. He found a change in the floret shape from
—
i=)
masson fossils of Berriochloa in
the oldest to the youngest stratigraphic levels: cy-
lindrical or evlindrical-fusiform shapes were found
in the oldest levels, while the youngest levels pre-
sented all possible forms within the genus: eylin-
drical. evlindrical-fusiform, obovoid. as well as
spheroid florets. The present study supports Tho-
masson’s findings, showing that obconical, орохо.
and lens-shaped florets are always derived states
(Fig. 3).
Species with obconical florets form a
316
Annals of the
Missouri Botanical Garden
Л,
М |
W
==,
mae. —
eS —
ڪڪ
а
LAE
Florets of North American species of Piptoc haetium. — P. avenacioides (Fredholm 5725). Note that the
D. P
Е
1. P. virescens (King & Soderstrom 5151).
Figure 5
scale of this drawing is 2 mm. —B. P avenaceum (Harvey
pringlei (Pringle 1410). —E. P. angustifolium (Машаа 29776 a). LE 2 seleri (Nee & Diggs 24775).
Pringle 8595). —H. P. brevicalyx (Rzedowski 35863).
G. P. fimbriatum
—
Volume 89, Number 3
2002
Cialdella & Giussan
Phylogenetic Relationships of Piptochaetium
monophyletic group derived from species with cy-
lindrical florets. Species with lens-shaped florets,
P. montevidense and P. panicoides, are derived spe-
cies within the Obovoid group (Fig. 3). The pres-
ence of papillae is another autapomorphy within
this small monophyletic group of lens-shaped flo-
rets, which is only present in living grasses and
never found in fossil taxa studied in this group of
Stipeae (Thomasson, 1985).
Piptochaetium angustifolium, P. fimbriatum, and
P. seleri (Fig. 5E, G, and F, respectively) have both
plesiomorphic (persistent awns and cylindrical flo-
м
rets) and apomorphic (short and blunt callus) states.
As shown by the phylogenetic analyses, these three
North and Central American species are intermedi-
ate in the evolutionary trend through the Obovoid
group.
Distribution patterns of species in Piptochaetium
suggest that these species found new and different
habitats to colonize in South America, with а con-
sequent higher diversification than in the Northern
Hemisphere. The monophyletic Obovoid group is
mostly South American, except P. brevicalyx and P.
montevidense, which are also present in Mexico, and
P. setosum, and P. uru-
guense, North and
South America (Fig. 4). According to Thomasson’s
suggestions (1980), disjunct species were first wide-
spread along animal migration routes and, later, due
P. stipoides var. stipoides,
with disjunct distributions ir
=]
to severe climatic fluctuation in the late Tertiary,
their distribution became gradually reduced, result-
ing in relict islands in both hemispheres. Whether
species with obovoid florets are derived from species
with cylindrical florets in North America or in South
America is difficult to determine, as there is no cer-
tain evidence for this fact.
INFRASPECIFIC LEVELS
Results from parsimony analyses showed a differ-
ent origin for subspecies of P. tovarii, first recognized
by Sánchez Vega (1991). Both subspecies of P. to-
varii are separated in all cladistic analyses (see
NONA and Pee-Wee results, Fig. 3). erigió
tovarii subsp. tovarii, included herein within the
burkartianum to P. panicoides clade. is characterized
by the absence of prickles on the lemma surface.
while P. tovarii var. pilosa (= P. pilosum herein) is
the sister group of the P. uruguense to P. stipoides
var. echinulatum clade and does have prickles and
macrohairs. Based on these results, the study of type
material, and the distribution of characters across
the entire genus, there is no reason to keep these
entities within the same species. Therefore, Р tovaru
subsp. pilosum is elevated to the rank of species (see
the taxonomic treatment herein).
Piptochaetium stipoides is included in the P. cu-
cullatum to P. stipoides var. echinulatum clade (Fig.
1).
stipoides were not resolved as much as among the
Relationships between the two varieties of P.
—
other species of the clade. A possible source of
variation within this clade is the polymorphic state
of P. stipoides var. stipoides for several characters
principally related with the floret. Moreover, tran-
sitional shapes and epidermal characters of the
lemma may be variable within a single specimen
of this species.
TAXONOMIC TREATMENT
Piptochaetium J. Presl, Reliq. Haenk. 1: 222.
830, nom. cons. Urachne Trin. sect. Pipto-
chaetium (J. Presl) Trin. € Rupr.. Mém. Acad.
Imp. Sci. St. Pétersbourg sér. 6, Sci. Nat. 5:
22. 1842. TYPE: Piptochaetium setifolium J.
Presl [= P. panicoides (Lam.) E. Desv.].
Podopogon Raf., Neogenyton: 4. 1825. nom. rejic. TYPE:
Stipa avenacea L. (lectotype, designated by Clayton,
м
1983, not seen)
Caryoc — Spreng., Syst. Veg. 4(2): 22. 30. 1827,
leg.. non Campanian Trinius. 826. TYPE:
yochloa montevidensis Spreng.
hom.
Car-
Caespitose, perennial plants. Culms herbaceous,
unbranched above, erect or somewhat decumbent,
terete or slightly flattened, longitudinally striate, gla-
brous; young shoots intravaginal. Nodes compressed,
occasionally thickened, the basal nodes slightly ge-
niculate. Leaf sheaths generally longer than the in-
ternodes at the base, becoming shorter toward the
apex. Ligules membranous, blunt, usually unfringed.
Leaf blades linear, convolute to flat, longitudinally
striate, frequently glabrous. /nflorescence a panicle
with branches densely to loosely flowered. Spikelets
fusiform, laterally compressed to terete, usually long
pedicellate, the pedicels unequal in length, generally
in pairs, rarely alone toward the apex of the panicle.
Flowers cleistogamous or chasmogamous. Glumes
two, persistent, subequal, usually longer than the flo-
et, membranous, lanceolate, acuminate, 3- to 7(8)-
nerved, glabrous. Florets easily deciduous, terete to
aterally compressed. Lemma indurate, awned, lon-
gitudinally striate, generally with epidermal append-
ages: lemma margins involute, fitting into the lon-
eitudinal groove of the palea; lemma apex fused in
a crown, frequently contracted to the base of the awn
and with different epidermal appendages. Awn usu-
ally twice-geniculate, twisted and hispid on the basal
югпоп. straight and shortly scabrous toward the
—
apex, deciduous or persistent. Callus acute, subacute
or blunt and obliquely truncate, usually hairy, rarely
glabrous or subglabrous. Palea laterally compressed,
318
Annals of the
Missouri Botanical Garden
indurate, membranous toward margins, glabrous,
shiny and inconspicuously striate, boat-shaped, bi-
keeled, the apex of the keels projecting as a minute
point above the summit of the lemma. Lodicules 2 or
3, membranous, blunt or acute, glabrous. Androeci-
um of З stamens always with short filaments, ca. 0.2
mm long, and small anthers, ca. 0.5 mm long in the
KEY TO THE SPECIES OF PIPTOCHAETIUM
cleistogamous flowers, and longer anthers 1-5 mm
long in the chasmogamous ones. Ovary obovoid, glo-
bose to laterally compressed, glabrous, with 2 styles
free to their bases; stigmas 2. Caryopsis terete to
globose or lens-shaped, surface dull, rugose and gla-
brous, with linear hilum and a small embryo; en-
dosperm hard, without lipid.
la. Florets terete, fusiform or obconical, occasionally fusiform, slightly obovoid; awn robust and persistent; callus
usualy acute or subacute, rarely blunt
emma glabrous
3b.
Lemma with inconspicuous ho
Lemma with hooks and prickles
за. F
ib.
Lemma With hooks all over surface, decreasing in density toward the callus and the —
ле with hooks and pric — on | the upper F ponini,
oks
10. P. cabrerai
— 25. P. palustre
always conspic uot
lo rel terete, not gibbous, slightly narrowed bi the crown, (б—)9—20 mm long, 1—
—
ume 7(8)-nerved; glumes 20-26(—35) mm long; lemma surface yr with
onspicuous longitudinal striae -
» P. hackelit
6b. ie glume 3(4)-nerved; glumes 15-23 mm long; lemma surface ЫЫ with
conspicuous longitudinal striae
Ta. Floret (14-)15-20 mm * RC 5. P. avenacioides
7 Floret n ge Tae )mmlong |... 24. P. napostaense
5b. Floret — slightly gibbous, or obconical, strongly narrowed below the crown,
(3.5-)7
8a. ie гаа,
Ja.
^
2
ОЬ.
етта;
— obe onical.
8b.
(—12) mm long, (0.8—)1.2-2 mm diam.
pus et 1.2-1.5 mm wide; lemma with hooks and prickles only below the
1; callus 3.5—4.8 mm long, occasionally shorter
Floret | mm wide; lemma with hooks and prickles on the distal half of the
callus 2-3 mm long
P. confusum
— . avenaceum
own with ascending macrohairs up to 0.6 mm * conspicuously ex-
pem the crown in length. sometimes also with a
=
—
llb.
oo
10b. бый йш macrohairs, with coke and puciles up to 0.2 mm
— tly — than
Crown 0.5-4
2 mm wide
12b.
кю
1За. "ua
is sharp and nar
a. Callus 1-2 mm long; floret (3.5—
I
C rown 0.8-1 mm diam.:
de
— with macrohairs — — all o over surfac e.
w hooks and p ee
)4.5-6(—7) mm long > bicolor
©
Callus 2.5—4 mm long; floret (5.5-)7.5-9.5 mm long —
D
ruprec —
long,
crown
vs mm dans 4 callus 1.2-1.5 mm long; floret (1.2—)1.
22. P. —
‘callus 1.9-2 mm long; floret 1.2-1.5 mm
7. P. brachyspermum
nfk orescence 3-9. em long, with 4 to 15 spikelets; spikelets 9713.5 mm long; Bolivia and
Arg
14b. Попе ‘ence longer, 10-30 cm long, with 15 to 80 spikelets; spikele ts 5-10 mm long.
PREND DERE P. indutum
Lower glume shorter than the floret; upper — as long as the floret; Brazil
15b. Both — — than the flore
l6a. Plant
spike
; Peru
— 1. P. alpinum
250.65 m long; — les 8—11.5 em long; panicles with 15 to 22
NR IA 30.
Д у ш
3
16b. Plant ^ 70-1. 30 m Jong; peduncles 40-44 cm ae panic les with 2 25 to 80
13b. Callus subacute or paie always broad.
l 7a.
) mm long:
Flo ret:
long; callu
ets; Mexico, Guatemala, and Venezuela
Lemma — sometimes a very few hooks present near the crown.
18а. К 7 callus subacute
is blunt
Р. virescens
. 28. P pringlei
15. Р in
17b. Lemma longitudinally striate, with hooks and pric ев all over surface
19а. Macrohairs of the lemma, including those of the crown, 1 mm е floret fusiform
' angu solum
19b. Macrohairs of the * тта, inc йш аве of the crown, 05 5 mm ions: floret terete,
slightly obovoid.
Volume 89, Number 3 Cialdella & Giussani 319
2002 Phylogenetic Relationships of Piptochaetium
20a. Peduncles 5-14 em long; inflorescences 3-6.5 cm long, dense: Peru -
14. p featherstonet
20b. Pedunc — 22-26 cm бан dali 'scences 57-220 cm i long. loose, not dense: н xico
and Guatemala: e ôç . P. seleri
Florets obovoid or lens-shaped: awn weak and deciduous; callus blunt.
21а. Lemma with many macrohairs
22a. Macrohairs of the елй up to 4 mm long, conspicuously longer than the crown 20. Р lasianthum
22b. Macrohairs of the lemma up to 2 mm long, the upper hairs slishily longer than the crown
— — 27. P. pilosum
—
21b. Lemma она,
23a. Aw n * eanie-baped, base as wide as the crown (0.9-1 mm), glabrous, or occasionally with very fev
habs. САРЕ GARE ASA 3. P. cuc —
23b. en thread-like. shortly hispid, minutely scabrous toward the ap
2 ‘rown 1-1.8 mm diam., not contracted toward the base of e awn.
25a. Callus glabrous. or with few short hairs as long as the callus.
26a. Hooks of the lemma inconspic uously pointed, eel disposed all over the
surface, decreasing in size and quantity toward the callus and near =й crown
, и сеп
м
26b. Lemma with very few licel. these confined to the distal part ian ig CTOW
e d › le —
25b. Callus densely hairy, the hans xe 'eeding the callus.
27a. Hairs of the callus longer than the floret 19. P. jubatum
27b. Hairs s the callus one hind or half as long as the floret.
28a. Lemma with pric ‘kles partially covered by the edges of — ent epidermal
s, generally in groups of 2 or 3, unifor mly disposed all over the lemma
surface, the biggest ones on dua айай portion of the floret
— 33b. P. stipoides var. echinulatum
|
e.
28b. — with hooks and pric ке not үш by the adjacent cells, uni
ormly disposed on the distal half of the floret, or lemma — epider ا
ndages sans 33a. P. stipoides var. stipoides
24b. C rown 0.4—0. 6-0. 9 mm diam., contracted toward the base of the awn.
a. Lemma with hooks ni wickles, densely and uniformly disposed on the distal % of
the floret; papillae absen
30a. Floret globose, sometimes slightly compressed laterally, (1.6—)2.4—3(-3.5) mm
long, 0.8-1.2 mm diam.; Paraguay, Uruguay, and northern Argentina
"——— P ee 35. P —
30b. Floret е and conspicuously Тами аренда, | (2 5-)3.8-4-5) 1
long, 1-2 1 ide; Се 2. и nse
29b. Lemma ib pay illae or with hooks and prickles only below the crown, or lemma
without epidermal appendages
3la. Crown revolute жану. the outside |... 2 83a. P. stipoides var. stipoides
31b. Crown not revolute.
32a. Floret lens-shaped, conspie ‘uously laterally compressed: lemma and crown
with or without papillae
33a. Lemma with — dense and uniformly — р рн, in
ul
size and density toward the crown and the callus . P. montevidense
33b. Lond without papillae, or confined to the region near the crown,
on the dorsal side of the floret 26. P. panicoides
32b. Floret globose to slightly compressed, — г obconical; lemma and
crown with or аш hooks, prickles, and macrohi р.
34a. Lemma with a few hooks toward the crown . — 82. P. setosum
34b. Le m oce epidermal appendages.
35a. Leaf blades glabrous: floret 2.2-2.5 mm long: callus hairy. the
upper hairs % the length of the floret › burkartianum
35b. Leaf blades hispid on both surfaces, sometimes aba Шу gla-
brous: floret 2.5-3.5 mm long; callus with or without short hairs,
the hairs equaling ilis le meth of the callus.
36a. Floret obovoid, not gibbous, 0.8-1.2 mm —
— . P. tovarit
36b. Floret conspicuously — 1.5-1.9 mm —
Glumes exceec ding t the length of the floret: lemma
dull, with conspicuous еи ıl striae; callus
0.4—0.5 mm long. crown with ] or 2 rows of hooks
and prickles on margin . P. hirtum
37b. Glumes as long as or slightly shorter than da floret:
lemma shiny, "amos callus 0.150.25 mm long:
crown without epidermal appendages _ 8. P. brevicalyx
320
Annals of the
Missouri Botanical Garden
Phyto-
Santa Ca-
B. Smith,
Brazil.
1. Piptochaetium alpinum L.
logia 22: 89. 1971. TYPE:
tarina: Bom Jardim da Serra, Fazenda da La-
ranja, 1400 m, Reitz & Klein 7710 (holotype,
US not seen; isotype, HBR not seen).
Distribution. Southern Brazil in the states of
Rio Grande do Sul and Santa Catarina, up to 1400
m elevation.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
2. Piptochaetium angolense Philippi, Anales
Univ. Chile 93: 734. 1896. TYPE: Chile. “An-
gol, Nov. 1887," Philippi s.n. (holotype, SGO
057391!).
Distribution. Endemic to central Chile (IV Re-
gión).
For a description, illustrations, апа specimens
examined, see Cialdella and Arriaga (1998).
90
.
Piptochaetium angustifolium (Hitchcock)
Valencia & Costas, Bol. Soc. Argent. Bot. 12:
177. 1968. Stipa angustifolia Hitchc., Contr.
U.S. Natl. Herb. ». 1925. TYPE: Mex-
ico. Coahuila: on summit slope of Sierra de L:
Puebla in sight of Saltillo, 25 July 1905,
Palmer 726 (holotype, US 570290 not seen:
isotypes, BAA Herb. Parodi 2955!, MO!, US
993385 not seen, US 906342 not seen
ures 2D, 5E
pau)
—
Fig-
Plants perennial. Culms 10-30 em tall, glabrous:
nodes 2, compressed, glabrous to sparingly pubes-
cent; internodes 3.5—4.5 cm long. Leaf sheaths
tightly embracing the culms, shorter than the in-
1-2 cm
long. blunt, margins entire, glabrous. Leaf blades
linear, convolute, 9-10(-20) em long, 0.5-0.6 mm
wide, longitudinally striate, glabrous. Peduncles te-
ternodes, 3—4 cm long, glabrous. Ligules
rete, slightly flattened, 7-10 cm long, longitudinally
striate, glabrous. Inflorescences 6.5-7.5 cm long,
few-flowered (8 to 18 spikelets),
loosely disposed; pedicels 2-7 mm long, scabrous.
with branches
Spikelets fusiform, 6—6.5 mm long, 1 mm diam.
Glumes longer than the floret, subequal, 6—6.5 mm
greenish, hya-
4.5-5
long, 0.9-1 mm diam. Lemma 4.5-5 mm long, con-
tracted
densely pilose, with macrohairs 1 mm long, hooks
long, shortly acuminale, 5-nerved,
line toward the margins. Floret fusiform, mm
below the crown, longitudinally striate,
and prickles all over surface; crown contracted at
the base of the awn, 0.4-0.5 mm diam., straight,
not revolute, with densely disposed prickles and
macrohairs, the latter | mm long, persistent; callus
0.6-1 mm long, subacute, densely pilose, the upper
hairs slightly longer than the callus; awn bigeni-
culate to almost straight, 10-11 mm long, persis-
lent, scabrous along length. Palea 4 mm long. Lod-
icules 2, 1 mm long, blunt. Caryopsis not seen.
Distribution, This species is known only from
northeastern and eastern Mexico, in the states of
Coahuila, Nuevo León, and Mexico, up to 3500 m.
куу, specimens examined. MEXICO, México:
elon de Nado, 2700-3500 m. Maruda e et D 29776
um. Nuevo León: Hac inda Jista Herm 35 mi. 5
of Monterrey, 810 m, 25 June 1939, Hanes 1038 (US).
Piptochaetium angustifolium is similar to P
featherstonei and P. seleri, but these species differ
in having short, 0.5 mm, macrohairs on the lemma
and crown, and a terete, slightly obovoid floret.
4. Piptochaetium avenaceum (L.) Parodi, Re-
vista Mus. La Plata, Secc. Bot. 6(25): 229.
1944. Stipa avenacea L., Sp. Pl. 1: 78. 1753.
TYPE: U.S.A. Virginia: Clayton 621 (lecto-
type, designated by Clayton (1983), LINN-
94.5 not seen; isolectotype, BM not seen). Fig-
ures 2C, 5
Stipa barbata Michx., Bor.-Amer. 1: 53. 1803,
ie g. non Stipa barbata — Fl. At
Stipa virginica Pers., Syn. Pl. 1: 99,
Papen barbatus Ral. « ex as D. Jac ks.. Index Kew.
. 1894, nom. inval., as syn. of Stipa avenacea
E m U.S.A “In sylvis Virginiae ( «arolinae,” Mi-
1. (holotype, P not seen; isotype, US not
nom.
er
Stipa а avenacea var. bicolor Eaton & J. Wright, Man. Bot.
(e d. 8): 444. 1848. TYPE: “United States of Amer-
ica” (not locate ad).
Stipa teiaha Hitche., Contr. U.S. Natl. Herb. 24: 236.
925. Piptochaetium и (Hitche.) Beetle,
4. 1983. TYPE:
: hytologia 54: l Mexico. Puebla:
2600 m, Hitchcock EA (holotype, US 993443 not
seen).
Piptoc — avenac es Barkworth, Syst. Bot. * 3: 196.
1988. TYPE: Mexico. erc ар 4 km W of on
bes 23% 2 N, №, on ime — itle
open pine forest, 31 0 ‚ 4 Aug. 1 Stanford e el
al. 645A n US | iol seen; vata es MO not
—
Plants perennial. Culms (30—)50—80 cm tall, gla-
brous; nodes 2-3, yellowish to reddish, com-
pressed, glabrous; internodes 5-25 cm long. Leaf
sheaths embracing the culms, shorter than the in-
(4—)6—-16 cm long, glabrous. Ligules 2-3
mm long, subacute to blunt, margins entire, gla-
ternodes,
brous. Leaf blades linear, convolute, occasionally
flat, 8-30 cm long, 0.8-0.9 mm wide, adaxial side
conspicuously striate, glabrous or scabrous, mar-
gins shortly scabrous. Peduncles terete, slightly flat-
tened, 8-10 cm long, longitudinally striate, gla-
Volume 89, Number 3
2002
Cialdella & Giussani 321
Phylogenetic Relationships of Piptochaetium
brous. Inflorescences 15-22 cm long, few-flowered
(10 to 15225) spikelets), with branches loosely dis-
posed; pedicels flattened,
Spikelets fusiform, 10—12
Glumes subequal, longer than the floret,
1.5-5 em long. hispid.
mm long, 1-1.2 mm diam.
10-12(—
15) mm long, acute, lower glume 3(5)-nerved, up-
per glume 5-nerved. Floret terete-fusiform, slightly
compressed, (7—)9-12(—13) mm long, 1 mm diam.
Lemma contracted below the crown, 9-12(-13) mm
long, thinly striate, with hooks and prickles toward
the distal portion (1/3-2/3 of the surface): crown
contracted at the base of the awn, straight, not rev-
olute, 0.5-0.6 mm diam., with prickles and macro-
hairs up to 0.5 mm long, occasionally up to | mm
long; callus ca. 2 mm long, sharp, hairy, the upper
macrohairs reaching 4-4 the length of the floret:
awn bigeniculate, 5-7 cm long, persistent, hispid,
scabrous toward the apex. Palea 7-8 mm long. Lod-
icules 2, 1.5 mm long, acute. Caryopsis terete, 5.5—
6 mm long, 0.5-0.6 mm diam.; hilum linear: em-
brvo Y, the length of the carvopsis.
Chromosome numbers. 2n = 22, 28 (Gould,
1958, as Stipa avenacea L.); n = 11 (Valencia €
1968).
Distribution.
Costas,
Widespread in the eastern and
southern United States and northeastern Mexico.
Alabama:
Additional specimens examined. —
ng Vani: 240,
Marshall Co.. margin of a woodland, а
about ТО mi. E of Morgan City, Hen der — Il~ (MQ): Lee
Co., Auburn, Earle & — 8. 73). ES insas:
Garland Co., 30 Apr. 1939, Den maree ped (US); Coastal
Plain Renan E maree 63483 (MO
Haven Co., dry
Weatherby 5
Wildlife
. Connecticut: New
yi PE es toward im of ML. Carmel,
Florida: Wakulla Co.. Marks
А СК 67795 е Co..
\ of Sebring. . Deam 64413 ( ; Gaine 8-
ue iS 99248 ,eorgia: Ben n Py
16.3 mi. ENE d Fitzgerald. Fokel 5079 (MO): Glynn
o., N end of Jekyll Island, Strong 1342 (US); Walker Co.,
near summit of ridge near Maddox Gap, between Villanow
and La Fayette; Cronquist 4432 (US). Indiana: N side of
the Pidgeon River about 3 mi. SE of Mongo, Deam 40701
(US). Louisiana: Pipeline off La. 507, 3 mi. W of Bien-
ville, Larrick 249 (ВАА); roadside about 2 mi. NE of
Mansfield, — 28558 (1 ы” т aryland: region of Marl-
: 12616 (US). sachusetts: Barnstable
C 20.. Hoffmann s.n. (MO): гн КО ul re
, near junction with hwy. 138, Blue Hill Re
ervation, SE of Boston. piri & Reeder zo (US). Mie h-
igan: 6 mi. W of Richland at Spring Brook and railroad,
Harvey 1764 (Sl); Kalamazoo Co.. Sect. 36, 2 mi. S of
Alamo. — 1854 (ВАА. US). Mississippi: Lowndes
j W of Mayhew junction on ridge, Me т 1811
a Co.. Bucatunna. Jacob 815 (US). New Jer-
ear Je kins, > Barrens.
Harkin 809 (МО); we 7312 (US). New
‘ork: Eastport, Long 1. land, Chase 7401 (US). North
Caroli na: Wake Co., Lake Johnson, 4 mi. SW of Raleigh,
— 3686 (US); Durham Co., wooded slope. Simpson
s.n. 2479778). Oklahoma: Le Flore Co..
near state
line on hwy. 63, E
Pennsylvania: s.l.,
Providence Co., Collins s.n. (
1 of US
of Big Cedar, Waterfall 14854 (US).
Scribner 266 (US). Rhode Island:
us oe South Caro-
. just SE of Ora. Миў
$ of Wale Be "ach; Griscom 16409
ina: Laurens Co., side
23032 (MO);
(US): MeCormick Co.. lr f Savannah River on SC
hwy. 28, Hos eman & Rin 8642 (US). Tennessee:
Marion Co.. 9 mi. SE of Tra City,
slo pe of ч wns Plateau, W of Rock-
Horry Сох
Roane Co..
wood, Wesker 6234 xas: Jasper Co., 8 mi. SE
of Jasper, hw 90, E н 7029 (SI); Shelby Co., к
scrub oak 3 7 mi. S of Center. Correll 16170 (US).
Virgi ia: picta Co., Chincoteague Island, шаи
8537 ( x Hampton, dry open woods, Miller Jr. s.n $
lT We уппа: on wooded — near Rey-
mon Memorial farm. Wardensville, Berkle ) (MO).
MEXICO. Tamaulipas: 3 mi. N of i p бй Hare
et al. 2482 (US, paratype of Piptochaetium avenacellum).
Piptochaetium avenaceum resembles P. confusum
and P napostaense, as they share the shape of the
floret and the appendages (prickles and hooks) on
the lemma and crown, but they are distinguished
as follows: in P. avenaceum, the prickles and hooks
are distributed on the distal portion. sometimes
reaching half the length of the floret. In P? confusum
and P. napostaense such appendages occur only im-
mediately below the crown. Both P. avenaceum and
P. napostaense have florets that are | mm in di-
but the
avenaceum and
ameter, callus length is 2-3 mm long in P
3.56 mm in P napostaense. Pip-
tochaetium confusum has florets 1.2-1.5 mm in di-
ameter with a callus 3.5-4.8 mm long.
5. Fiptochadtium avenacioides (Nash)
Costas, Bol. Soc. Argent.
Stipa y ere
. Club 22
T Lake on bi pine dn near Cassia, 16/
30 June 1895, Nash 2051 (holotype, NY not
BAA!, MO!, US!). Figure 5A.
Valencia
Nash.
TY PE: U.S.A. Flor-
seen: isotypes,
Plants perennial. Culms 0.70-1.20 m tall, gla-
brous or pubescent below the basal node ы nodes
yellowish. glabrous: internodes long.
Leaf sheaths ¢ embracing the culms, shorter than the
Ligules 1-3
glabrous.
—
internodes, em long. glabrous.
(-2
mm long, blunt or acute, margins entire,
Leaf blades linear, convolute, 15-30 em long. | mm
wide, longitudinally striate, glabrous. Peduncles te-
22-25 cm
brous. Inflorescences 10-30 cm long, few- to many-
rete, long. longitudinally striate, gla-
flowered (10 to 50 spikelets), with branches loosely
disposed: pedicels 1.5-2 em long. scabrous. Spike-
lets
Glumes longer than the floret, 18-22 mm long, acu-
fusiform, 18-22 mm long, 1-2 mm diam.
minate, (3)5-nerved. Floret terete, (14-)15-20 mm
long, 1-1.2 mm diam. Lemma 15-20 mm long, con-
tracted below the crown, longitudinally striate, with
322
Annals of the
Missouri Botanical Garden
small hooks on the distal portion (1/3-2/3 of the
surface); crown contracted at the base of the awn,
0.6-0.7 mm diam.,
crohairs 0.4—0.5
mm long,
with hooks, prickles, and ma-
mm long, persistent; callus 6—8
acute, densely hairy; awn bigeniculate,
persistent, 8-12 mm long, hispid at the basal por-
tion, scabrous toward the apex. Palea 12 mm long.
Lodicules 2, ca. 2.5 mm long, acute. Caryopsis not
seen.
Distribution and ecology. Only known from the
United States, in Florida. This species grows in dry
and sandy places.
Additional specime ns. "red ed. U.S.A. Florida: Bre-
vard Co., Fredholm 5725 (BAA, MO, US); Eau Gallie,
Curtiss 5834 (MO, US); н Co., Mermiack, Baker 115
US); Manatel Co., Palma Sola, Tracy 7031 (MO, 05); Du-
val Co., Fredholm 5092 (US); W of Tavares, Hitchcock 812
(US); Braidentown, Hitchcock 967 (US); s.l.. Dowell 7 7309
(BAA, US); Orange Co., Clarcona, Meislahn 191 (US
—
6. E bicolor (Vahl) E. Desvaux, Fl.
Chil. E 1853. Stipa bicolor Vahl, Symb.
Bot. — "id Oryzopsis bicolor (Vahl)
Speg.. "iia us. Nac. Montevideo 4(2): 6.
1901. Piptoch — bicolor var. typicum Pa-
Mus. La Plata 6(25): 252. 1944,
. TYPE: Uruguay. “Habitat in Bra-
siliae Mons Video. Dn. Thouin," Herb. Vahl,
IDC microfiche photo Vahl, nr. 73 Ш, 4-5 (ho-
lotype, С!).
Stipa — Trin. & Rupr.,
étersbourg sér. 6, Sci. Nat
zl. "Brasilia me топа — 9 784 (holotype,
LE not 3 not see d
Stipa Lt Sag ke Steud., Syn. Pl. Glumac. 1: 124. 1854.
* Chile. Valparaiso, Concon, Nox 1829, Herb.
s pri >
Oryzopsis bicolor (Vahl) S
jus Acad. Imp. Sci. St.
26. 1842. TYPE: Bra-
seen; isotypes, US!,
“ minor Speg., Anales
. 1901. Piptoc от
us. Nac. onte — ^0 102 y <
bicolor (Vahl n . Desv. var. minor opa g.) Parodi, Re-
vista Mus. La Plata, Secc. Bot. 6(25): 256. 1944.
TYPE: Argentina. Buenos Aires: Bavio, Estancia
Elizalde, Spe gaz ie s.n. — LPS 12513! sheet
A; isotype, LPS 12513! sheet В).
Distribution. Central Chile, in V Región, west-
ern and southern Uruguay in the departments of
Canelones, Florida, Montevideo, and San José, and
central eastern. Argentina in the provinces of Bue-
nos Aires and Entre Ríos.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
7. Piptochaetium brachyspermum (Spegazzini)
Parodi, Revista Mus. La Plata, Secc. Bot.
6(25): 241. 1944. Oryzopsis napostaensis Speg.
var. brachysperma Speg., Anales Mus. Nac.
Montevideo 4(2): 17. 1901. TYPE: Argentina.
Buenos Aires: Carmen de Patagones, Spegaz-
zini 42 b (holotype, LPS 2471!)
Distribution. Endemic to Argentina, where it is
only known from the southern province of Buenos
Aires.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
8. Piptochaetium brevicalyx (E. Fournier) Ric-
er, Contr. U.S. Natl. Herb. 17: 286. 191
Mexic. Pl. 2: 150.
=
Stipa brevicalyx E. Fourn.,
1886. TYPE: Mexico. San Luis Potosf: cerca
de San Luis Potosí, 6000-8000 ft., 1878, Par-
ry & Palmer 959 (holotype, US not seen; iso-
types, BAA!, K not seen, MO not seen, NY not
seen). Figures 1A, 2A, 5H
Piptochaetium brevicalyx (E. Fourn.) Ricker subsp. flexuo-
um Barkworth, Syst. Bot. = 200. 1988. TYPE:
i of Durango City on
road to La Flor, 2620 m ale, Breedlove 44215
(holotype, CAS 622475 not seen).
Culms (6—)10—50 cm tall;
nodes 3 or 4, compressed, reddish, glabrous; inter-
Plants perennial.
nodes 2-12 cm long. Leaf sheaths tightly embracing
the culms, shorter than the internodes, 2-5 cm
long, glabrous to scabrous. Ligules 0.6-1.2 mm
ong, blunt to subacute, margins entire or fringed,
glabrous. Leaf blades linear, convolute or flat, 5—30
cm long, 0.6-1.2 mm wide, glabrous or hispid. Pe-
duncles terete, 2.5—7 cm long, longitudinally striate,
glabrous. /nflorescences 3—10 cm long, few-flowered
(5 to 20 spikelets), with branches densely disposed;
pedicels = mm long, glabrous or hispid. Spikelets
globose, 2.5-3.5 mm long, 1.5-1.7 mm diam.
Glumes subequal, shortly acuminate, violet-colored,
hyaline toward the margins, 5- or 7-nerved; lower
2.4-3.3 mm
long: upper glume shorter than or as long as the
floret, 2.4—3.5
compressed, gibbous, 2.5-3.5 mm long, 1.5-1.9
glume slightly shorter than the floret,
mm long. Floret globose, slightly
mm wide. Lemma contracted below the crown, 2.5—
3.5 mm long, smooth, shiny, without epidermal ap-
pendages: crown contracted at the base of the awn,
inconspicuous, straight, not revolute, 0.5-0.7 mm
diam., without epidermal appendages; callus 0.15-
0.25 mm long, blunt, obliquely truncate, with few,
occasionally many, short hairs; awn slightly genic-
ulate to almost straight, 6-18 mm long, hispid, sca-
brous toward the apex, deciduous. Palea 2.5-3.7
mm long. Lodicules 2, 1 mm long, acute. Caryopsis
not seen.
Distribution.
tely 3000 m.
Central Mexico, up to aproxima-
Additional specimens examined. MEXICO. Durango:
5.3 km W of (Ejido) Los Mimbres turnoff, litis et al. 214
(US); 41 km NE of El Salto, Mick & Roe 87 (US! paratype
Volume 89, Number 3
2002
Cialdella & Giussani 323
Phylogenetic Relationships of Piptochaetium
of P. brevicalyx subsp. flexuosum). Guanajuato: 10 km al
sur de Ibarra. Ocampo, sobre la carretera a León, Rze-
dowski 50768 (MEXU). Hidalgo: 8 km al noreste de Pa-
ica, sobre la carretera a Real del Monte, Rzedowski
(MEXU, XAL): Cerro Ventoso. above Pachuca.
Pringle 7606 (MEXU): in oak woods near Real del Monte,
Rose et al. 8680 sack Pachuca, Hitchcock 6733 (US).
México: rocky roadcut, 27 mi. S of Juan del e
Gould 9211 (US 5). San pus Potosí: Cerro La Bolsa,
' Cerrito s Dolores, Villa de Arriaga. B el
di 132 (CAS, MO): near Cerro Prieto, Sohns 1535 (US).
Tlaxcala: js. San Cristóbal and Calpulalpan. Sole
553 (MEXU. US).
Piptochaetium brevicalyx is similar to P. hirtum:
both species share the globose floret shape. the ab-
sence of epidermal appendages on the lemma, and
generally few hairs on the callus. They can be rec-
ognized because P hirtum presents conspicuous
longitudinal striae on the dull lemma surface. | or
rows of hooks or prickles in the crown, and both
glumes 5.5-6 mm, longer than the floret.
Barkworth (1988) recognized Piptochaetium bre-
vicalyx subsp. flexuosum based on the relatively
long and widespread branches of the inflorescenc-
es. Although there are specimens with dense or ex-
tended panicles, the difference between the length
of the branches is slight. and, within the studied
material, there are specimens with both types of
inflorescences, such as Banda et al. 132
*
ону burkartianum Parodi, Revis-
ta Mus. lata, Secc. Bot. 6(25): 291. 1944.
TYPE: пан Corrientes: San Martin, La
Cruz. Nov. 1936, Parodi 12329 1/2 (holotype.
BAAS).
Distribution. Endemic to Argentina; only found
in the eastern province of Corrientes, at the type
locality.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
10. Piptochaetium cabrerae Parodi, Revista
Mus. La Plata. Secc. Bot. 6(25): 243. 1944.
TYPE: Argentina. Buenos Aires: Puan, Villa
Iris. 7 Nov. 1940, Parodi 13771 (holotype,
BAAN.
Distribution. Endemic to Argentina: only col-
lected in the southwestern province of Buenos Ai-
res.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
11. Piptochaetium calvescens Parodi. Revista
Mus. La Plata. Sece. Bot. 6(25): 278. 1944.
TYPE: Argentina. Prov. Buenos Aires: Sierra
Currumalán, 600 m s.m., 10 Nov. 1932, Pa-
rodi 10343 (holotype, BAAN.
Distribution. Central and eastern Argentina in
the province of Buenos Aires, and southern Uru-
guay in the department of Montevideo.
‘or a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998)
12. Piptochaetium confusum Parodi, Revista
Mus. La Plata, Secce. Bot. 6(25): 246. 1944.
TYPE: Argentina. Entre Ríos: Concordia. 3
Nov. 1921, Parodi 3950 (holotype. BAA!).
Distribution and ecology. Uruguay in the de-
partments of Maldonado, Montevideo, and Paysan-
dú, and central eastern Argentina in the province
of Entre Ríos. This species grows in dry grasslands
and rocky places.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
13. Piptochaetium cucullatum Rosengurtt &
= Univ. Rep. Fac.
‚ 1960. TYPE: Uru-
guay. Dpto. Salto: — del Arapey. próximo
9 Dec. 1962,
МУКА not
US 2951786 not seen).
Izaguirre de Arturio,
Agron. Montevideo 90:
a la Estación,
1512 (holotype,
BAA!,
Arrillaga et al.
seen: 180ty pes.
Distribution. Endemic to northern Uruguay in
the departments of Salto and Paysandú.
‘or a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998)
(Hitchcock)
14. Piptochaetium featherstonei
Tovar, Opusc. Bot. Pharm. Complut. 4: 104.
1988. Stipa featherstonet Hitche., Proc. Biol.
Soc. Wash. 36: 196, 1923. TYPE: Peru. Rio
Blanco, 4500 m, 8-19 May 1922. J. Е Mac-
bride & W. Featherstone 803a (holotype, F
517 í 331
seen).
pro parte not seen: isotype, US not
Piptoc ноне Мети Tovar & Gutte,
egni Veg. 91(4): 205. 1980. TYPE:
auli, La Oroya, Berg oberhalb
Pachacayo (Puya н Vorkommen). Steiniger
Abhang, Hohe, 3750 m. Gutte 2130 (holotype. LZ
not seen: vendi SME not seen, USM!).
Feddes Repert.
Distribution. Southwestern Peru in the depart-
ments of Ancash, Ayacucho, and Junín at eleva-
tions between 3000 and 4500 m.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
Vote. Hitchcock (1923) that the
holotype is mounted with plants of Stipa hans-
mentioned
meyeri, but, according to Soreng (pers. comm.). it
is not clear if that specimen was subdivided into
parts a and b by the collectors or by Hitchcock
324
Annals of the
Missouri Botanical Garden
himself; the US isotype has only Stipa feathersto-
net on it.
15. со fimbriatum (Kunth) Hitch-
cock, J. Wash. Acad. Sei. 23(10): 453. 1933.
Spa pu Kunth, Nov. Gen. Sp. 1: 126.
1816. Oryzopsis fimbriata ee Hemsl.,
Biol. Cent.-Amer., Bot. 1885. TYPE:
Mexico. Guanaxuato, Billalpando, Humboldt &
Bonpland s.n., Herb. HBK 4224 (holotype, P
not seen; isotype, BAA!). Figures 2B, 5G.
Piptoc — — (Kunth) ү s Eo —
M. Johnst., J. A : : Ў JO.
г! — ‘bor . TYPE:
Mexico. ca uila: 'a del jm » La n. ‚ John-
ston & Mueller 186 үкөм GH по! seen; —
LL not seen).
Plants perennial. Culms 35-80 cm tall, glabrous,
sometimes pubescent below the nodes; nodes 2—3,
slightly compressed or enlarged, reddish, glabrous;
internodes 6-11 em long. Leaf sheaths tightly em-
bracing the culms, shorter than the internodes,
(3—)5-8 cm long, glabrous. Ligules 1.5-2 mm long,
acute, glabrous. Leaf blades linear, convolute or flat,
6-26 cm long, 0.9-1.1 mm wide, adaxial side con-
spicuously striate, glabrous on both faces, margins
scabrous. Peduncles terete, 12—28 cm long, longi-
tudinally striate, glabrous. /nflorescences 14-20 cm
long, 60 spikelets), with
branches loosely disposed; pedicels flattened, 4—12
many-flowered (20 to
mm long, hispid. Spikelets fusiform to slightly ob-
ovoid, 3.5—5 mm long. | mm diam. Glumes sube-
qual, longer than the floret, 4-6 mm long, shortly
acuminate, 5- or 7-nerved. Floret fusiform, slightly
obovoid and gibbous, 3.5-5 mm long, 1.2-1.8 mm
diam. Lemma contracted below the crown, 3.5—5
mm long.
hairs, densely disposed; crown contracted at the
smooth, shiny, with deciduous macro-
base of the awn, inconspicuous, straight, not revo-
lute, ca. 0.8 mm diam.,
ages; callus 0.2-0.5 mm long,
truncate, hairy, the upper macrohairs reaching V,
the length of the floret; awn bigeniculate, 10—20
mm long, persistent, hispid to scabrous. Palea 3.5—
3.6 mm long. Lodicules 2, ca. | mm long, subacute.
Caryopsis fusiform, 2.5 mm long, 0.6 mm diam.;
the length of the cary-
without epidermal append-
blunt, obliquely
hilum linear; embryo 1/5
opsis.
Chromosome numbers. 2n
n = 21 (Reeder, 1968); 2n =
tas, 1968).
Distribution,
southeastern United States and Mexico, at eleva-
tions between 1900 and 2500 m. This species has
— 44 (Brown, 1951):
44 (Valencia & Cos-
Widespread in the southern and
also been cited for Guatemala, in the department
of Huehuetenango (Swallen, 1955; Beetle, 1977).
ap iud — ns examined. U.S саве
Сапе elo Hills, along me 15 ni. N of
& McHargue 2222 (MO); Patagonia
ey € Peobles 10114 (US); Cochise Co.,
Creek Chiricahua Mountains, Peterson & Annable 1010
— Lh New M v.
104,
ex
©
Y
722
© М
= №
355
cky
\ . Correll 13099 (US
tains, Brewster Co., l mi. SW of Boot Spring, 2120 m,
Moore & Steyermark 3172 (MO, US). MEXICO. Baja Cal-
Sur: granitic — surrounding long interior val-
ley (La Laguna), S of Pico La Aguja on the Sierra La
Laguna, Breedlove & — 43293 (MO). Chiapas: Teo-
=
Chihuahua: foothills
MEXU); Miñacam, Hitchcock 7741 (US
navista, Saltillo, Cañón San — nzo. 5
cia Zacatecas, Madrigal A. s.n. (XAL); Canon San Loren-
zo, 5 km de Saltillo hacia Zac 'atecas, Serrato Sánchez s.n.
(MEXU 309021). Distrito Federal: Cima de Toluca,
Lyonnet 48 (MO). e on top of rugged volcanic
plateau 5 5 »3 km W of (Ejido) Los Vibes turnoff al km
54.2 WSW of — and ca. halfw El Salto, /ltis
el al. 213 (XAL). Guanajuato: ca. 8 km W of San Felipe,
Sohns 424 (MO, US). Hidalgo: open pine woods near
Ocotillos between — 0: and Huasca, ca. 2200 m,
Moore & Wood Jr. 4164 ). ТТ о: Mezquitic, 8 km
SE del Rancho El Mortero, — ski 17704 (US). Méx-
ico: Tezer oco, lado sur de la Cañada de Aguas, 13.5 km
s E de Tezcoco. 11 km al ESE de Coatlinchán, Koch
77110 (MEXU); — old hwy. 190 between turnoff to
c haleo (Hwy. 115 ita Bárbara ca. 30 m above Azotla,
] . Mie оаа 0.6 mi. E of Poblado
ca. 20 mi. W of Morelia,
. Nuevo Leon: 28.6 mi. N La Ascen-
Brunken & Perino 234 (MO). Oaxaca:
7 km N de быш. Nochixtlán, Mendoza is Pia 2640
MO). Puebla: Tepeyahualco, Calzada et al. 4645 (XAL);
vicinity of San Luis Tultilanapa, Purpus 3589 (US). San
uis Potosí: Cerro La — 2.5 km S de La Ama-
pola, Escalerillas, Zavala Ch. 862 (XAL); in canyons in
the Sierra de San Miguelito, ca. : dus ы of Te srrero, 1850—
2200 m, 8 Sep. 1954, Sohns 1147 (U
de los Ajos, dro ho de los Ajo т уа
et al, 92-794 (MO). Tamaulipas: Marcella. Stanford et al.
2604 (MO). Tlaxcala: between San Cristóbal and Capu-
lalpan in open, lightly wooded areas, Sohns 587 (MEXU)
Contadero, Pringle 8595 (MEXU). Zacatecas: 9.5 mi. W
of Sombrerete, Taylor & Taylor 6254 (US)
—
Ф 5
m
ES
oc
N
EX
pu
The epidermal surfaces of the lemma in Pipto-
chaetium fimbriatum and P. pringlei are similar, as
both species present smooth and shiny floret sur-
faces with the macrohairs deciduous. However, P.
pringlei differs in having bigger florets, from 7 to
10 mm long, and a subacute callus.
Volume 89, Number 3
2002
Cialdella & Giussani 325
Phylogenetic Relationships of Piptochaetium
16. Piptochaetium hackelii (Arechavaleta) Pa-
. Revista Fac. 7(1): 162.
Anales Mus.
1895.
Anales
4(2): 10. 1901. Pipto-
Agron. Veterin.
Stipa hackelii Arechav.,
Nac. Hist. Nat.
Oryzopsis hackelii (Arechav.) Speg.,
Mus. Nac.
Buenos Aires 4: 179
Montevideo
chaetium hackelii (Arechav.) Herter. Estud.
Bot. Reg. Urug. 4: Florula Urug.: 34. 1930,
comb. superfl. TYPE: Uruguay. Cerro de Mon-
Arechavaleta 39
isotype, LPS 2480!
tevideo, cerca de la cumbre.
a (holotype, not located:
US not seen).
Stipa — Kuntze, Rev. Gen. Pl. 3(2): 373. 1898.
TYP Nov. 1892,
Argentina. Buenos Aires: Tandil.
Otto Pea s.n. (holotype, NY!
Distribution. This species grows in the grass-
lands of Uruguay in the departments of Lavalleja.
Maldonado.
Argentina in the provinces of Buenos Aires and La
and Río Negro. as well as in central
Pampa.
For a — illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
Note The holotype of Stipa hackelii Arechav.
could not be located. According to Alonso Paz
(pers. comm.), Arechavaleta did not choose types.
so in many cases a lectotype must be designated.
17. — ‘haetium hirtum Philippi. Anales
Univ. Chile 43: 559. 1873. TYPE: Chile. Cerro
Dec. 1864. Philippi s.n. (lectotype.
designated by Cialdella & Arriaga (1998).
SGO 45078!; isolectotypes, K!, SGO 57422!).
de Renca,
— haetium З Phil. Anales Univ. Chile 93:
33. 1896. — Cerro de Renca, 11 Nov.
1877. 20 s.n. (le . designated by Cial-
della & Arriaga (1998), pr 0 45077!;
SGO 45078!, SGO 57421).
isolec lolvpes.
Distribution. Known from Cerro de Renca it
Chile and San Martin de los Andes in the province
of Neuquen in Argentina (Parodi, 1961).
For a description and illustrations, see Cialdella
and Arriaga (1998).
18. rapes tium indutum — Revista
lus. La Plata, Sece. Bot. 6(25): 258. 1944.
TYPE: Argentina. Salta: Rosario de Lerma.
Puerta Tastil, 2700 m, Venturi 8414 (holotype.
US not seen: isotype, BAAD.
Distribution. Originally known from northern
Argentina from the provinces of Salta and Jujuy.
distributed in Bolivia in the departments of La Paz
and Potosí, central western Peru in the departments
2500
it is also cited from Ecuador in the
La Libertad, and Lima. between
1500 m:
of Ancash,
and
province León-Y ánez.
19991.
of Azuay (Jorgensen «
PERU. Ancash: Re-
Río Pachacoto, mw &
. Huascarán National
Additional specimens артлап.
cuay Prov., Huascaran irk,
Torres 11794 ; (MO); Tie Prov
Park. Llanganuco Sector. Smith 10488 (MO): Recuay
Prov., Quebrada Huanca, Smith: & Buddensiek 10964
(MO). La Libertad: Sánchez
machuco road, 20 km
quez M. 3327 (MO
W of и ену ан Smith & Vás-
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998
19. Piptochaetium jubatum Henrard, Recueil
Trav. Bot. Néerl. 36: 537. 1939. TYPE: Uru-
guay. Canelones: Río Santa Lucía. Paso Cue-
По. Gallinal et al. 2198 (holotype, Herb. Lugd.
Bat. n. 938,280-383, 11:
1723425 not seen, BAA!)
isotypes, US
Distribution. Known only from the department
of Canelones in Uruguay.
For a
examined,
description, illustrations, and specimens
see Cialdella and Arriaga (1998).
20. Piptochaetium lasianthum — Grisebach,
Symb. Fl. Argent.: 297. 1879. TYPE: Argen-
Entre Ríos:
Quinta del Colegio, en praderas. Lorentz s.n..
tina. Concepción del Uruguay.
Flora Entrerriana n. 1157 (holotype. not lo-
CORD).
cated; isotype,
— s erianthum Balansa. Bull. Soc. Bot. France
i і. 1885. TYPE: Uruguay. Cerro de Montevi-
s "0. “187 4. Balansa 9 (holotype. not located).
—
Distribution. Southern Brazil in the states of
Rio Grande do Sul and Santa Catarina, central and
southern Uruguay in the departments of Flores, La-
valleja, and Montevideo, and widespread in central
eastern Argentina in the provinces of Buenos Aires,
Córdoba. Corrientes, Entre Rios, Misiones. and
Santa Fe.
Additional specimen examined. ARGENTINA. Misio-
nes: Candelaria. Villa Venecia, Renvoize et al. 3025 (MO).
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998).
21. Piptochaetium la. (Spegazzini)
Henrard, Recueil Trav. Bot. Néerl. 36: 530.
1939. Oryzopsis — Speg., Anales Mus.
Nac. Montevideo 4(2): 19. 1901. TYPE: Ar-
gentina. Sierra de la Ventana,
Valle de Nov. 1895, C. L. Spe-
баат s.n. LPS 12660).
Buenos Aires:
las Vertientes,
(holotype.
Distribution. Southern Uruguay in the depart-
© A
326
Annals of the
Missouri Botanical Garden
ments of Lavalleja and Maldonado, and central
eastern Argentina in southern Buenos Aires.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
22. Piptochaetium medium (Spegazzini) Torres,
Bol. Soc. Argent. Bot. 11(4): 251. 1969. Ory-
zopsis bicolor (Vahl) Speg. var. media Speg..
Anales Mus. Nac. Montevideo 4(2): 9. 1901.
TYPE: Argentina. Buenos Aires: Sierra de Cu-
ramalal, Dec. ‚ C. L. Spegazzini s.n. (lec
totype, designated by — (1969), LPS
12517!).
Distribution. Southern Brazil in the state of Rio
Grande do Sul, southern Uruguay in the department
of Florida, and eastern Argentina, widespread in
i Córdoba, Entre
Ríos, La Pampa, Misiones, and Santa
the provinces of Buenos Aires,
Additional specimens examined. BRAZIL. Rio
Grande do Sul: Esmeralda, 29 Nov. 1988, M. Sallés 141,
142 (MO)
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998).
23. Piptochaetium montevidense (Sprengel)
Parodi, Revista Fac. Agron. Veterin. 7(1): 163.
1930. Caryochloa montevidensis Spreng., Syst.
eg. 4(2): 30. 1827. Oryzopsis montevidensis
(Spreng.) Hauman, p us. Nac. Hist.
Nat. Buenos Aires 29: 116. 1917.
montevidensis (Spreng.) Speg., Revista Argent.
Bot. 1(1):
montevidensis (Spreng.) Speg. f. typica Speg..
Revista Argent. Bot. 1(1): 11. 1925,
Piptochaetium montevidense (Spreng.
Estud. Bot. Reg. Urug. 4: Florula
І 930, comb. superfl. TYPE: Uru-
guay. Montevideo: Sellow s.n. (holotype, B not
seen; isotype, MO!)
Oryzopsts
1925, comb. superfl. Oryzopsis
nom. in-
—
Urachne panicoides Trin. var. brasiliensis Trin. & Rupr.,
— Acad. — Sci. St.
dis t. 5: 23.
ar. basili ensis (Trin. & Rupr.)
Bot. 13(10): 40. 1913. TYPE
ad fines regni Paraguayani,”
10t located).
Piptoc haetium tuberculatum E. Desv., in Gay, Fl. Chil. 6:
272. 1 ryzopsis tuberculata (E. Desv.) Speg.,
Anales Mus. Nac. Montevideo 4(2): 26. 1901. TYPE:
Chile. Valdivia: Gay s.n. (holotype, P!).
в — Steud., Syn. Pl. Glumac. 1: 123. 1854.
Chile. Rancagua: Bertero 456 (holotype, P!).
Piptoc * ie verrucosum Phil., Linnaea 33: 280. 1865.
zopsis verruculosa (Phil.) Speg., error for verru-
Anales Mus. Nac. Montevideo 4(2): 28. 1901.
Oryzopsis verrucosa (Phil.) Speg., Revista Argent.
Pétersbourg sér. 6, Sci.
342. Caryochloa montevidensis Spreng.
) Doll ex Ekman, Ark.
"A Montevideo usque
° Sellow s.n. (holotype,
cosa,
Bot. 1(1): 12. 1925. TYPE: Chile. Valdivia: San
Juan, — 5. з fnoletype, SGO 57334'; probable
isotype, SGC 350!).
ae и бите Phil., Anales Univ. Chile 93: 730.
TYPE: Chile. colo on: médanos, Philipp s.n.
е 6 45087
Piptocl haetium diiit Phil, Anales Univ. Chile 93:
732. 1896. TYPE: Chile. Concon: Philippi s.n. (ho-
lo otype, SGO heats
ШЕ moelleri Phil., Anales Univ. Chile 93: 734.
896. TYPE: Chile. Renaico, Möller s.n. iu
7332.
) 578:
Oryz zopsis montevidensis (Spreng.) Speg. f. оо
.. Revista Argent. Bot. 101): 11. 1925. TYPE
Argentina Misiones: Posadas, praderas de * alre-
dedores, Spegazzini s.n. (holotype, LPS 13166!
PENNE montevidensis (Spre ng) Speg. f. ase
levista Argent. Bot. з ): 11. 1925. ТҮРЕ:
Brasil Rio Grande do Sul: Uruguayana, campos se-
cos, Spegazzini s.n. eee LPS 13175).
Distribution. Found in Venezuela in the state of
Trujillo, and widespread in southern South America
in Bolivia, departments of La Paz and Cochabamba,
central Chile, IV, VI, and УШ Regions, southern
Paraguay, departments of Misiones and Paraguarf,
southern Brazil, states of Paraná, Rio Grande do
Sul, and Santa Catarina, southern Uruguay, depart-
ments of Colonia, Lavalleja, and Maldonado, and
northern and central Argentina, provinces of Bue-
nos Aires, Chaco, Cérdoba, Corrientes, Entre Rfos,
Misiones, Salta, and Santa Fe. This species, which
has been considered disjunct in Mexico, northern
South America, and southern South America (Tho-
masson, 1980), has also been cited for Peru (Brako
& Zarucchi, 1993) and southern Ecuador, province
of Chinchipe (Jørgensen € León-Yánez, 1999).
This distribution pattern makes it probable that it
also occurs in Colombia, and, together with P. pa-
nicoides, would confirm one of the possible migra-
tion routes proposed by Thomasson (1980).
This species is the most frequent of the genus,
growing on riversides, grasslands with rocky places
and modified soils, up to 3700 m elevation.
Additional specimens examined. BRAZIL. Paraná:
Ruta BR- 280, hacia — nte, a 1 km de ruta BR-153,
Rúgolo et al. 1611 (MO). VENEZUELA. Trujillo: Cara-
che, 220 m, Rivero & Díaz 1384 (MO).
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998).
24. Piptochaetium A dup n (Spegazzini)
Hackel, Anales Mus. Nac. Hist. Nat.
Aires 11: 103, 1904. Ts napostaensis
. Anales Mus. Nac. Montevideo 4(2): 15
. Oryzopsis napostaensis Speg. var. ma-
crophylla Speg., Anales Mus. Nac. Montevideo
4(2): 17. 1901. TYPE: Argentina. Buenos Ai-
res: alrededores de Carmen de Patagones, Feb.
1898, C. L. Spegazzini 41 a (lectotype, desig-
nated by Parodi (1944), LPS 2483!).
Buenos
Volume 89, Number 3
2002
Cialdella & Giussani 3
Phylogenetic Relationships of Piptochaetium
Oryzopsis napostaensis Speg. var. — Speg., Ana-
les Mus. Nac. Montevideo 4(2): 17. 1901. TYPE:
Arge ntina. Бас Aires: Sierra de Мр Dec.
899 . Spegazzini 41 A (holotype, s 69!).
Stipa — Hack., Anales Mus. Nac ie
nos Aires 11: 95 1904. TYPE: presea е
Dep. Rio Т, — Villamonte, Fa.
— 13803 type. nol seen; Isosynlypes,
ORD!, LPS!), aie 14082 (syntype, not located).
Stipa ton Hack., Aires
13. 1911. TYPE; Pu Buenos
ación он ЕС... M. Estrada s.n.,
kert 17397 Бе ada W not
3168626 not
San Teodoro,
Anales Mus. Nac. Bu
Hab. Siue:
isotype, US
seen;
seen).
Distribution. Endemic to central Argentina in
the provinces of Buenos Aires, Catamarca, Córdo-
ba, La Pampa, Mendoza, Río Negro, and San Luis,
up to 1900-2000 m elevation. This species grows
in dry and arid grasslands and is a frequent species
of the
tree” (Prosopis caldenia Burkart), in the
“caldenal.” an open wild forest of “calden
“Espinal”
biogeographic province.
‘or a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
25. Piptochaetium зки Mujica-Sallés &
Longhi-Wagner, Candollea 48: 15. 1993.
TYPE: Brazil. Santa Catarina: Urupema, junto
Retransmisora del Morro de
Campo Novo, 1680 m s.m., 25 Nov. 1984, J.
F. M. Valls et al. 8083 (holotype, ICN!; isotype,
CEN.
la Estación
Distribution. Known only from southern Brazil
at the type locality.
For a description and illustrations, see Cialdella
and Arriaga (1998).
26. Piptochaetium panicoides (Lamarck) E
Chil. 6: 270. 1853. Stipa
. Genr. 1: 158. 1791.
Speg.,
‚ 1901.
without locality,
Desvaux, in Gay, Fl.
panicoides Lam.,
Anales
—
Oryz — panicoides (Lam.
Mus. Nac. Montevideo 4(2): 31 TYPE:
Uruguay. Montevideo: 1767.
Commerson s.n. (isotypes, P!, MPU not seen,
US not seen).
Hist. Nat. 12: 445. 1827.
Oryz > — A. Rich., Dict.
: (holo-
: Uruguav. Montevideo: iba TSON S.H.
type, probably P not seen).
Piptoc — setifolium J. Presl, Reliq. Haenk. 1: 222.
1830. Stipa vil (J. Presl) — وا PL 1:
182. br 833. Urachne simplex Trin. & Rupr.
e
=
ruviana Trin. К ao m
Pétersbourg sér 842. Ory ZODsIS
setifolia (J. P a He — — ш rb. 40:
57. 1921. TYPE: Peru. Sin localidad, T. Haenke s.n.
(holot . PR).
Oryzopsis e arpa Speg., Anales Mus. Nac. Montevideo
42) 33.
1901. Piptochaetium leiocarpum (Speg.)
Hack., Anales Mus. Nac. s Nat. Buenos Aires
13: 463. 1906. TYPE: Arge . Tucumán:
пава, Lorentz & Hieronymus 608 ( (holotype.
‘ated; isotype, LPS!).
— — — Phil., Anales Univ. Chile
1896. * : Chile. Río Itata, — s.n. D
m SG * not seen; isotypes, SGO 57374!, US not
een).
iu — leiocarpum (Speg.) Hack.
دغ
a
ilé-
not lo-
f. Up want
lack. ex Stuck., Anales Mus. Nac. Hist. . Bue-
nos Aires 13: 463. 1906. Piptochaetium „л,
(Lam.) E. Desv К. ex Stuck.)
f. subpapillosum (Hac
Parodi, Revista Mus. La Plata, Secc. Bot. 6(25): 302.
1944. Piptochaetium ралбай (Lam.) E. Desv. var.
Miri AMAA (Hack. ex Stuck.) Petetin.
FI. ónica, Colecc. Ci. Inst. Nac. Tecr
: Argentina.
without locality, Lillo 4314 (h ody pe not located;
isotype, Herb. Arg. Stuckert 15418, CORD!).
Stipa — Mez, Feddes Repert. Spec. Regni Veg.
. 1921. TYPE: Argentina. Tucumán: La Cié-
naga. — & Hieronymus 608 (holotype. B not
seen).
Oryzopsis pis arpa Speg. var. major Speg., Revista Argent.
Bot. 1(1): 10. 1925. TYPE: n нв, Jujuy: Sierra
de Santa Barbara, Sta. Spegazzini 2490
(осете: LPS 126411).
еј
Tucumán:
Cornelia,
Distribution. This species has a disjunct distri-
bution within South America: an Andean area, in-
cluding mountain regions of Venezuela, Colombia.
and Peru (Brako & Zarucchi, 1993), widespread in
Ecuador (Jørgensen € León-Yánez, 1999), Bolivia.
—
Chile, and Argentina, in the provinces of Córdoba.
Jujuy, Salta, and Tucumán, and a coastal zone lim-
ited to sandy soils of southern Brazil, Uruguay, and
Argentina, in the provinces of Buenos Aires and
Entre Ríos.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
27. Piptochaetium pilosum (Sánchez Vega)
Cialdella & Giussani, stat. nov. Basionym: Pip-
tochaetium tovarii Sánchez Vega subsp. pilosa
Sánchez Vega, Arnaldoa 1(1): 29. 1991. TYPE:
Peru. Cajamarca: entre Llacanora y Namora,
2750 m, Sánchez Vega et al. 2459 (holotype.
CPUN!: isotypes, MO!, USM!
The florets of this species have lemmas with
prickles, hooks, and macrohairs all over the sur-
face, and a callus that is densely pubescent. These
characters clearly distinguish Piptochaetium pilo-
sum from Р, tovarti, with florets without any epi-
dermal appendage on the lemma and callus. These
differences, together with the results of the cladistic
analysis, make it reasonable to raise Piptochaetium
tovarii subsp. pilosa to the rank of species.
Distribution. Peru in the departments of Caja-
marca and Junín.
328
Annals of the
Missouri Botanical Garden
For a description and illustrations, see Cialdella
and Arriaga (1998
28. Piptochaetium pringlei (Beal) Parodi, Re-
vista Mus. Plata, Secc. Bot. 6(25): 230.
1944. Oryzopsis pringlei Beal, Bot. Gaz. 15:
112. 1890. Stipa pringlei (Beal) Scribn., in Va-
sey, Contr. U.S. Natl. Herb. 3(1): 54. 1892.
TYPE: Mexico. Chihuahua: dry ledges, Sierra
Madre, 5 Nov. 1887, 8500 ft., Pringle 1410
(holotype, US!; isotype, MO!). Figure 5D.
Plants perennial. Culms 0.5—1 m tall, terete, gla-
brous, pubescent below the nodes; nodes 2—3, red-
dish, compressed or enlarged, glabrous or slightly
pubescent; internodes 7-20 cm long. Leaf sheaths
embracing the culms, shorter than the internodes,
6-18 ст long, scabrous, hispid to the base. Ligules
1-2.5 mm long, blunt or subacute, margins entire,
hispid at the abaxial side. Leaf blades linear, flat,
10-30 cm long, 1-3 mm wide, glabrous, scabrous
on the nerves and margins. Peduncles terete, slight-
ly flattened, 30-50 cm long, longitudinally striate,
hispid. Inflorescences 15-16 ст long, few-flowered
(10 to 25 spikelets), with branches
posed; pedicels flattened, up to 1 mm long, hispid.
Spikelets fusiform, 9-12 mm long, 1.2-1.9
diam. Glumes subequal, longer than the floret, 9—
loosely dis-
mm
12 mm long, acuminate, glabrous; lower glume 5-
or 7-nerved; upper glume 7-nerved. Floret terete-
fusiform, 7-10 mm long, 1—1.5(-1.9) mm diam.
Lemma contracted below the crown, 7-10 mm long,
smooth, shiny, with yellowish macrohairs, decidu-
ous, and few prickles toward the crown; crown con-
tracted at the base of the awn, inconspicuous,
straight, not revolute, 0.4-0.8 mm diam..
of long macrohairs 0.8-1 mm long and prickles;
callus 0.8-1 mm long, subacute, obliquely trun-
cate, hairy, the upper hairs slightly longer than the
callus; awn bigeniculate 1-2
(—3.5) em long, persistent, hispid, scabrous toward
with rows
to almost. straight,
the apex. Palea 6-7 mm long. Lodicules 2, «
mm long, acute. Caryopsis not seen.
Chromosome numbers. 2n = 42 (Myers, 1947,
as Stipa pringlei (Beal) Scribn.); n = 21 (Reeder,
1977, as Stipa pringlei (Beal) Scribn.).
Distribution. Southwestern United States and
northern Mexico, up to 2000-2500 m elevation.
Additional specime ns examined. U.S.A. Arizona: Co-
arr Canyon, Huachuca Mountains,
Williams, Hitchcock. 1903
near Tucson, So-
Silveus 3470
— hcock
3 (US).
E “Cube T-
chise Co., 6 mi. up
Gould elt des 2456 (B. АА, US);
(US); Mt. Lemmon, Santa Catalina Mts.,
derstrom — (US). California: Santa Rita,
(TEX). New Mexic :0: Ranger Station, Queen,
813 (В. АА); Guadalupe Mountains, Hitchcock 1
Texas: Black Mountains, Tharp 4175 (TEX, l
son Co., S McKittrick Canyon, top of Guadalupe Mts.,
Warnock 12082 (TEX); Lincoln Co., Hinckl
Jr. 629 (US). Chihuahua: Sierra Madre reverie Sierra
Gazachic 35 km SW of Minaca, Pennell 18929 (US); Si-
erra La Viga, 6 km E de Jame, Puerto Maravillas, Villa-
rreal et al. 1975 (XAL). Sonora: Morelos, El Rancho de
Robles, Vera Santos 1959 (MO, US); E of Cananea, Sierra
de Los Ajos, Beetle M-7856 (MO).
—
Piptochaetium pringlei is slightly similar to P.
fimbriatum in that both species share the smooth
surface of the lemma, only with macrohairs and few
hooks near the crown, Piptochaetium fimbriatum
differs in having a shorter floret, 3.5—5 mm long,
and a blunt callus.
29. Piptochaetium ruprechtianum К. Des-
vaux, in Gay, Fl. Chil. 6: 274. 1853. Oryzopsis
ruprechtiana (E. Desv.) Speg., Anales Mus.
Nac. Montevideo 4(2): 12. 1901. Stipa ruprech-
tiana (E. Desv.) Herter, Revista Sudamer. Bot.
6(5—6): 141. 1940. TYPE: Without lo-
cality, Sellow s.n. (holotype, LE-Trin not seen;
Brazil.
isotype, US not seen; drawing of US isotype,
Herb. Parodi 4127 sheet b, BAAN).
Oryzopsis — (Vahl) Speg. var. major Speg., Anales
Mus. Nac. Montevideo 4(2): 9. 1901. TYPE:
tina. с. de Tandil, DAMM s.n. (holotype, LPS
2515!; isotype, BAA!
Argen-
Distribution, Southern Brazil in the states of
Rio Grande do Sul and Santa Catarina, Uruguay in
the departments of Lavalleja, Maldonado, and So-
riano, and Argentina in the provinces of Misiones
and Buenos Aires. This species is frequent in rocky
prairies, and in Brazil it grows in savannas.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
30. Piptochaetium sagasteguii Sánchez Vega,
Arnaldoa 1(1): 1991. TYPE:
marca: Cerro El Guitarrero, ladera occidental
del valle de Cajamarca, 2800 m, 15 Jan. 1983,
I. Sánchez Vega 2914 (holotype, CPUN!; iso-
types, AAU not seen, CHAPA not seen, F not
seen, HAO not seen, K not seen, MO!, US
3232467 not seen, USM not seen).
Peru. Caja-
Distribution.
ments of Cajamarca and La Libertad at elevations
between 2700 and 3500 m.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
Known from Peru in the depart-
Volume 89, Number 3
2002
Cialdella & Giussani 329
Phylogenetic Relationships of Piptochaetium
31. Piptochaetium seleri (Pilger) Henrard, Blu-
1940. Oryzopsis seleri Pilg., Verh.
192. 1909.
mea 3: 452
Bot. Vereins Prov. Brandenburg 51:
TYPE: Guatemala. Huehuetenango: between
Todos los Santos and Chiantla, 3000 m.
Seler 3238 (holotype, В not seen; isotypes,
BAAL! US 2767420 not seen). Figure 5F.
Sep. ?,
Plants perennial. Culms 20-90 em tall, terete,
glabrous; nodes 2 or 3, reddish, compressed, gla-
)10-15 ст
sheaths embracing the culms, generally shorter than
brous: internodes (4.5— long.
the internodes, 8-9 em long, glabrous. Ligules 0.8—
2 mm long. blunt to subacute, membranous, mar-
gins entire, glabrous. Leaf blades linear, convolute,
5.5-12 cm long. 0.4-0.5 mm wide, abaxial side
scabrous on the nerves, adaxial side glabrous. Pe-
duncles terete, 22-26 cm long, longitudinally stri-
ate, glabrous. Inflorescences 7—22 cm long. few to
many flowered (25 to 60 spikelets), with branches
loosely disposed: pedicels angled, 0.4—1.5 ст long.
glabrous. Spikelets fusiform, 4-5 mm long. 1-1.4
mm diam. Glumes subequal, longer than the floret.
shortly acuminate, 5-nerved, viola-
Floret
34.5 mm long. 1-1.2 mm diam.
1-5 mm long.
ceous, hyaline toward the margin. terete,
slightly obovoid.
Lemma narrowed below the crown, 34.5 mm long.
striate. with short macrohairs and hooks all over
the surface: crown contracted at the base of the
not revolute, 0.4—0.5 mm diam., with
callus 0.4-0.6 mm
hairy,
awn, straight,
prickles and short macrohairs;
long. blunt, obliquely truncate, the hairs
slightly longer than the callus; awn bigeniculate,
1-1.5 em long, persistent, shortly hispid. Palea ca.
3.5 mm long. Lodicules 2. 0.8-1 mm long. acute.
Caryopsis terete, slightly obovoid, 3-3.2 mm long.
0.8—0.9 mm diam.:
length of the carvopsis.
hilum linear: embryo %4 the
Widespread in Mexico and south-
between 2000 and 3600 m.
Distribution.
ern Guatemala,
Selected specimens examined. MEXICO. Coahuila:
Cañón de San Lorenzo, Sierra de Zapalinamé, 3 km S of
Saltillo, Snow et al. 6721 (MO). Distrito Federal: De-
sierto de los Leones, Lyonnet 2719 (US); lava fields ca. 2
km SSW of La Cima, R.R. station on — side of old
hwy. 95, 14 Aug. 1960, Ticis et al. 958 (US). Hidalgo:
woodlands near Trinidad Iron Works. eee 13249 (MO,
S) México: Sierra de Las Cru Pringle 5200
(МЕХ. US): 10 km al E de Amecameca. sol
tera a Tlamacas.
27 km SW of Toluca on road to Temaxcaltepic.
Roe 180 (US): 19202 (МЕХ):
Llano Grande, near Río Río. Sharp 44144 (US): Chapingo.
Ixtapaluca, faldas del Cerro El Papayo. Campo Experi-
mental Zoquiapán, Zavala Ch. 245 (XAL); al SW del Cen-
tro Experimental Zoquiapán, Velazco Torres & J. Trinidad
s.n. (XAL); Telapon, Lyonnet & Elcoro 1948 (US): Parque
Nacional Lagunas de Ze "mpoala, Traylor 61 (МЕХ, US).
ces,
re la carre-
Leaf
Laguna de San Gre-
san Andre
Beaman 4273. нени
EXU).
Santa Clara del Cobre,
nit of Cerro S
Michoacán: `
gorio, Escobedo 1476 (XAL): sumn
km N of Ciudad Hidalgo.
elos: Le mpoala. Lyonnet 2511 (
zi
>
ca. 12
1529 (US). Mor
Oaxaca: 18 mi. SW of city of Oaxaca, Nelson 1373
US). Tlaxcala: ca. 3 mi. NE of Tlaxco, v Sols 604 (U 5),
Veracruz: Acajete, Rzedowski 11940 б XU): Calca-
hualco, end of passable portion of road from i Cos -
pec-Esc ola-Jacal to Miguel Hidalgo and Tlac hichuca, Nee
& Diggs 24775 (MO, XAL); La Joya, Mejra S. 218 (XAL)
Las Minas, vereda Cruz dar anca a Rinconado. india
E. & Burgos 498 (XAL): igas,
pasar la presa del ij Fo rumbo a Te pes газ.
Chazaro & — 3792 (XAL)
Ventura A. 18506 (XAL): Sandoval. 68
AL); Perote. por la brecha que va a Tonalaco en las
faldas del Cofre de Perote. Castillo et al. 1980 (XAL):
Soledad Atzompa, Wer" gr * Acosta r 1063 (XAL); Xico.
Tonalaquillo, Arriaga C. AL). GUATEMALA —
huetenango: Sierra de vn a uc vri 8, . immediate ly
of Tojiah at km 322, on ruta nacional 9 392 Ў
(US); near Tunimá, Sierra de los Cue matan ss,
mark 48274 (US); trail between Soloma and 5
Sierra de los Cuchumatanes, Steyermark 484, 5 WS). To-
tonicapán: Desconsuelo, potrero nature flora alpina, de
Koninck 114 (US); on the Tecum Uman Ridge at km 154
ca. 20 km E of Totonicapán, Bea-
km despues de
Manzanares, Las
Perote. Escobillo.
“><
—
. Beaman
on ruta nacional nro. 1.
man 4155 (US).
Beetle (1977) and Hitchcock (1951. under Ory-
zopsis seleri) placed Piptochaetium seleri in synon-
ymy with P. fimbriatum. Both species have terete.
slightly obovoid florets and lemma wilh macrohairs.
but P. fimbriatum has smooth or thinly striate lem-
mas, without prickles or hooks, and a crown without
epidermal appendages.
32. Piptochaetium setosum (Trinitus) Arecha-
valeta, Anales Mus. Nac. Montevideo 1: 330.
1896. Urachne setosa Trin., Mém. Acad. Imp.
Sci. St. Pétersbourg sér. 6, Sci. Nat. 6(1): 124.
1834, nom illeg. superfl. TYPE: Chile. With-
out locality, Cumming s.n.. TRIN 1473.1 (lec-
totype. designated by Parodi (1944). LE not
seen: Isotype, US not seen).
е — dye ex Steud., Syn. Pl. Glum l: 123.
. TYPE: C E pt te — 153 (iso-
Iv pe. o
iare i purpuratum Phil., Linnaea 29: 1857
TY Chile. Valparaíso, Nov. 18547 — ^»
lotvpe. SGL 574191).
Piptoc haetium — ex Griseb., Symb. Fl. Ar-
зат TYPE: Chile. Cerro San С ristóbal,
"a اس
re
8
Philipp s.n. designated here, SGO
Anales Univ. Chile 93:
Pi iptoc Jn macrocarpum Phil..
896. TYP : А | V. Briones s.n. (ho-
(35. ). ше. Nuble,
lotype. SGO 45082!).
Distribution. This species presents, at the mo-
ment, a disjunct distribution in southern North
America (California) and South America (Bark-
330
Annals of the
Missouri Botanical Garden
worth, 1986, 1993), where it grows in central Chile
from Valparafso to Valdivia.
Observation. This species has a strikingly pros-
trate habit in some specimens collected in Califor-
nia (Barkworth, pers. comm.).
r a description, illustrations, and specimens
199
-rj
‘or
examined, see Cialdella and Arriaga
Note. Piptochaetium setosum (Trin.) Arechav.
was based on Urachne setosa Trin., a nomenclatur-
ally superfluous name when published, as Trinius
cited Stipa panicoides Lam. and Oryzopsis setacea
A. Rich. as synonyms, both earlier published. Nev-
ertheless, Р, setosum is the correct name in Pipto-
chaetium for Urachne setosa when Stipa panicoides
and Oryzopsis setacea are considered synonyms of
—
another species, Piptochaetium panicoides (Lam.
E. Desv
33. Piptochaetium stipoides (Trinius & Ru-
precht) Hackel, Anales Mus. Nac. Montevideo
1(4): 328. 1896.
33a. Piptochaetium stipoides (Trinius & Ru-
precht) Hackel var. stipoides. Urachne —
Mém. Acad. Imp. 8
Pétersbourg, sér. 6, Sci. Nat. 5: 25. 1842. ra
zopsis stipoides (Trin. & Rupr.) Speg., Anales
Mus. Nac. Montevideo 4(2): 23. 1901. Pipto-
chaetium stipoides (Trin. & Rupr.) Hack. var.
genuinum Parodi, Revista Mus. de La Plata,
des Trin. € Rupr.,
есе. Bot. 6(25) 266. 1944, nom. inval.
TYPE: Brazil. Brasilia meridionalis, Sellow
n. (holotype, LE not seen).
Symb. Fl. Argent.:
Piptoc 'haetium oe Griseb.,
298. 1879. Piptochaetium ov — Desv. var. chae-
Anales Mus. Nac. Hist.
е 1904. TYPE: Argentina.
Córdoba: Dpto. Punilla, in de Azúcar, Hieronymus
s.n., Herb. Grisebach m — GOET not seen;
isotype, CORD!).
„шш — — Phil., Anales Univ. Chile 93:
‚ 1896. TYPE: Chile. e. de Conchalí, Philippi
s.n. s holotype . SGO 57401!; isotype, US 1819510 not
о (Grise ij Hac
n).
Piptoc — ovatum E. Desv. var. purpurascens Hack.,
Anales Mus. Hist. Nat. Aires 21: 86.
1911. Piptoc — — Trin. & Rupr.) Hac o
var, n ens (Hack.) Parodi, Revista Mus. de
Plata, Secc. 6(25): 272. 1944. TYPE: Arge —
Chaco: — 1 de Mayo, Margarita — ‘n, Stuckert
19189 (lectotype, here ds 'signated, (
Piptoc — ovatum E. Desv. f. atrata Hu Anales
Mus. Nac. Hist. Nat. p поз Aires 21: 85. 1911.
ina. Chaco: Dpto. | de Mayo, Colonia
18304 (holotype, W not seen;
Buenos
TY РЕ: Argent
Benitez, Stuckert
type, CORD!).
Stipa verruculosa Mez, Feddes Repert. Spec. Nov. Regni
leg. 17: 206. 1921. Piptoc — — ulosum
(Mez) Henrard, Meded. Rijks-Herb. 380. 1927.
Piptochaetium verruculosum (Mez) i. Revista
iso-
Bot. 6(5-6): 141. 1940, comb. superfi.
Argentina. Buenos Aires: without locality,
Balansa s.n. (holotype, B not seen; isotype, US not
seen
Sudamer.
TYPE:
Piptochaetium — Gpeg. ) Herter, Revista Sudamer.
* e ). Oryzopsis E Speg..
us монеті 4(2): 4. . TYPE:
mentia: E ntre e Río : Concepción del * lay, Lo-
rentz s.n., a Entrerriana 1691 — С ORD!;
isotype, | Pst),
Pichi stipoides (Trin. & Rupr.) Hack. var. parvi-
florum Parodi, Revista Mus. de La Plata, Secc. Bot.
6( 75. 1944. TYPE: Argentina. Buenos Aires:
Pdo. Balcarce, Balcarce, Martinez Crovetto 950 1/2
(holotype, BAA!).
Distribution. This variety presents, at the mo-
ment, a disjunct distribution (Thomasson, 1980;
Barkworth, 1986, 1993) in the southwestern United
States and Mexico (Beetle, 1977), and in South
America, where it grows in southern Brazil in the
states of Rio Grande do Sul and Santa Catarina. It
is also widespread in Uruguay, in Canelones, Co-
lonia, Flores, Florida, Lavalleja, Montevideo, Pay-
sandú, Río Negro, Salto, Soriano, and Tacuarembó;
in Chile, in УШ and IX Regions and in Región
Metropolitana; and in Argentina in the provinces of
Buenos Aires, Catamarca, Chaco, Córdoba, Corri-
entes, Entre Ríos, Jujuy, Misiones, Río Negro, and
Santa Fe. This variety is known up to elevations of
2000 m in arid regions.
ARGENTINA. Cha-
Additional pm —
| Nov
co: Margarita én, оу J08, Stuckert 19208
(CORD). MEXIC О. Tamaulipas: 5i de San —
Armadillos, 10 July 1930. Bartlett. 1022:
Cerro de los
—
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998).
33b. Piptochaetium stipoides (Trinius & Ru-
precht) Hackel var. echinulatum Parodi, Re-
vista Mus. La Plata, Bot. 6(25): 271.
TYPE: Argentina. Buenos Aires: Pdo.
Villa Parodi 13759 (holotype,
Secc.
Puán, Iris,
ВАА!).
The florets of this variety present the lemma with
prickles partially covered by adjacent n s
Cialdella & Arriaga, 1998:
—
cells at the base
3D, 10J)
Distribution.
central Argentina in the provinces of Buenos Aires,
La Pampa, and San Luis, and in northern Uruguay
This variety has been found
in the department of Salto.
For a description, illustrations, and specimens
examined, see Cialdella and Arriaga (1998).
Volume 89, Number 3
2002
Cialdella & Giussani 331
Phylogenetic Relationships of Piptochaetium
34. Piptochaetium tovarii Sánchez Vega, Ar-
naldoa 1(1): 25. 1991. TYPE: Peru. Cajamar-
ca: Cajamarca, entre Cajamarca y Cumbema-
yo, Fundo Universidad, 3450 m, 22 May 1971,
I. Sánchez Vega & N. Vilhena 678 (holotype.
CPUN!: isotypes, MO!, US 3232468 not seen).
Distribution. Southern Ecuador in the limited
region between provinces of Loja and Azuay, and
northwestern Peru in the departments of Piura, Ca-
jamarca, and Ancash, at elevations between 2700
and 3450 m.
ECUADOR. Azuay-
Loja: Nudo de cordillera occidental y cordillera oriental,
entre Ofia y Rane 5 Ovejero, 1-2 Aug. 1959, Barclay &
Juajibioy 8455 (
Additional specimens examined.
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998).
35. Piptochaetium uruguense | Grisebach.
Symb. Fl. Argent.: 297.
guayensis (Griseb.) Speg., Anales Mus. Nac.
Montevideo 4(2): 29. 1901. Piptochaetium uru-
var. genuinum Parodi, Revista
1879. Oryzopsis uru-
guense Griseb.
Mus. La Plata. Sece. Bot. 6(25): 290. 1944.
nom. inval. TYPE: Argentina. Entre Ríos:
Dpto. Uruguay, Concepción del Uruguay. 13
Nov. 1875. P. G. Lorentz 471 (holotype, prob-
GOET not. seen; BA!
ably in isotypes,
CORD!).
— —— ш ксн var. microcarpum Parodi,
evista Mus. de ita, Secc. Bot. 0(25): : 290.
۳ М. TY "we ntina. godes ntes: Dpto. San Mar-
tín. s ú, Parodi 12637 (holotype, BAA!).
=
=
Distribution and ecology. For the time being.
this species presents a disjunct distribution in Mex-
1980; Barkworth, 1986) and
South America, in Paraguay, departments of Misio-
ico. (Thomasson,
nes and Guairá, in western Uruguay, departments
of Cerro Largo, Flores, Florida. Paysandú, and Ro-
cha, and in northern and northeastern Argentina,
provinces of Chaco, Corrientes, Entre Rios, Misio-
Santa Fe. Piptochaetium uruguense
nes, Salta. and *
also grows in southern Brazil, state of Rio Grande
do Sul, and it has been cited for
(Smith et al., '
ries.
Santa Catarina
1982). This species grows in wet prai-
Additional examined. BRAZIL. Rio
Grande do Sul: Fazenda E ns rimental de Criacao Bage.
| Dec. 1945, Swallen 7561 (MO)
spec ime n
For a description, illustrations, and other speci-
mens examined, see Cialdella and Arriaga (1998).
36. Piptochaetium virescens (Kunth) Parodi,
Revista Mus. La Plata, Secce. Bot. 6(25): 230.
1944. Stipa virescens Kunth, Nov. Gen. Sp. 1:
126. 1816. TYPE: Mexico. Guanajuato, Santa
Rosa, Cuesta de Belgrado y La Buffa, Hum-
boldt & Bonpland s.n. (holotype, P not seen:
isotype, US 00141712 not seen). Figure 5C.
—
Repert. Spec. Nov. Regni Veg. 8: 515
(Kunth) Parodi var.
Stipa arsenii Hack.,
1910. Piptochaetium virescens
arsenii (Hack.) Beetle, Phytologia 54: 4. 1983
TYPE: Mexico. Michoacán: Morelia, 2200 m. Fr. Ar-
sene 3211 (holotype. W not seen: US nol
MO 845899 not seen).
isotypes.
seen,
Plants perennial. Culms 0.70—1.30 m tall. terete,
glabrous, occasionally hispidous below the nodes:
nodes 3 or 4, brownish, compressed, glabrous: in-
ternodes 12—45
the culms, shorter
em long. Leaf sheaths embracing
than the internodes, 9.5-15 cm
long. glabrous. Ligules 1-2.5 mm long. blunt, mar-
gins entire, glabrous, sometimes hispidous at the
10-35
cm long, 0.5-1 mm wide, longitudinally striate, gla-
abaxial side. Leaf blades linear, convolute.
brous, scabrous at Ше margins. Peduncles terete,
40-44 em long, longitudinally ribbed, glabrous. /n-
florescences 10-30 cm long. many-flowered (25 to
80 spikelets), with branches loosely disposed: ped-
icels slightly angled, 2-18 mm long. hispid. Spike-
6-8 mm long. 1.2-1.4
Glumes subequal, longer than the floret, 6-8.5 mm
lets. fusiform, mm diam.
hyaline toward
&(—9) mm
long. | mm diam. Lemma slightly contracted below
5.5-8
hooks and macrohairs uniformly disposed: crown
long, shortly acuminate, 5-nerved,
the margins. Floret К ed 5.5—
the crown, mm long. thinly striate, with
contracted at the base of the awn, straight, not rev-
olute, 0.5-0.7 mm diam., with prickles and rows of
fa)
macrohairs 0.5-0.6 mm long. occasionally 1 mm
long; callus 0,9—1 mm long, acute. occasionally
subacute, densely hairy, the upper macrohairs
slightly longer than the callus; awn bigeniculate.
13-20 mm long, persistent, hispid, scabrous toward
the apex. Palea ca. 5 mm long. Lodicules 2. 1.4—
3—3.2 mm
embryo Y, the
1.6 mm long. acute. Caryopsts terete.
long. 0.9 mm diam.; hilum linear:
length of the caryopsis.
Chromosome numbers. 2n = ca. 60 (Gould,
1966, as Stipa virescens).
Distribution. Guatemala and central and south-
ern Mexico. This species was cited also for Vene-
zuela, Barinas. in TROPICOS based on J. Brunken
& C. Perino 355 (MO), which we did not see.
GUATEMALA. Hue-
huetenango: Cumbre Papal, between summit and La Li-
bertad, Steyermark 50955 (US). Quezaltenango: La Es-
Selected specimens examined.
Annals of the
Missouri Botanical Garden
peranza, De Koninck 73 A and B (US). San Marcos:
Barrancos 6 mi. S and W of town of Tajumulco, NW slopes
of Volcán Le жеты Steyermark 36700 (US). MEXICO.
Chiapas: 5 exican hwy. 190 near Rancho Nuevo,
about 9 mi. of San Cristóbal las Casas, Bre a
14171 (US). Distrito dense) е de Mexico, С. de
Ajusco, Matuda 2575: XU): S. de Ajusco, Pringle
6588 (MEXU, MO, US) ен У Serranfa de
Ajusco, Lyonnet 1877 (MO); 4 km W of San Andreas,
Pedregal de San Angel, Sohns 177 (US). sa Tepea-
pulco, Cerro de Xihuingo, Ventura . А. 376 (MO); Tulancin-
go, open pine woods near reservoir at — between
Acaxochitlan and Puebla, Moore Je 2843 (US); Real de
Monte, Matuda 18907 (US). Jalise — km SW of
Tequila on Volcán de Tequila, Век оге 39241 (MO); Za-
potlán, Hitchcock 7169 (US); Sierra del Tigre. 5 mi. S of
Colim:
de Mexico, D nac ho.
Amecameca, Matuda 25724 (ME т О); Ix-
aluca, 8 km S de Río Frío, Koch 762 (MO); Popo Park,
pr — 5965 (US); Cerro de Pinal, Otzoloapan, Matuda
el al. 31875 (US). Michoacán: in open pine forest anc
in and around the NE side of the Volcán de
Р aric utin, Sohns 819 (MEXU); ca. 18 mi. S of Patzcuaron,
ing & Soderstrom 5151 (MEXU): Las Cañas, Rzedowski
«€ McVaugh 623 (MO); Las Cañas, estribaciones inferiores
noroccidentales del Cerro Patamban, Tangancicuaro, Rze-
dowski & McVaugh 623 (US). Morelos: Sierra de Ajus
Pringle 62. * و XU, MO, US); Lagunas оен ж
пе! 2498 . Oaxaca: 23 km de Tlaxiaco, rumbo a
Chalcotongo, — M-4743 (MO); 0.5 km : de Las Huer-
tas, Nundichi, Manzanero M. 450 (M l En
al S de San Andrés, carr. Oaxac "a Puerto m
huatlan, Torres & Cedillo 875 (MO).
М de Tlaxco, al S de Cerro La Paila, Chimal et al. 9 (MO);
5 mi. NE of Zac 'atepec, on hwy. 40 from Jalapa, Soderstrom
pass, 46 mi.
n open areas
Tlaxiaco,
. to Mexico City,
(US). San Luis Potosi: on the NE slopes of hills near
Aguaje de García in the Sierra de Guadalcazar,
1500 (US). Tlaxcala: Huamantla, 500 m al SW
tamirano Guadalupe, Guerrero et al. 632 (MO); 3 mi. NE
of Tlaxco, Sohns 601, Sohns bis IS). Veracruz: Aya-
XN 228 (MO). Zacatecas: collected near
Plateado, Rose 2750 (US).
Sohns
Piptochaetium virescens is similar to P. indutum:
they share the floret and callus shape and the epi-
dermal appendages on the lemma. Piptochaetium
indutum can be recognized because it includes
smaller plants, 15-35 em tall, with smaller florets,
(4—)5-6.25
glumes, (7-)8-13.5 mm.
mm long and 1 mm diam., with longer
Piptochaetium virescens and P. sagasteguii are
also related species, with similar floret characters:
P. sagasteguii may be recognized because it in-
cludes smaller plants, 25-65 em tall, with few-flow-
ered inflorescences of 15 to 22 spikelets. The geo-
graphical distribution of these species is also
different: P. sagasteguii is only known for Peru,
while P. virescens can be found in Mexico and Gua-
temala.
EXCLUDED SPECIES FROM PIPTOCHAETIUM
Piptochaetium mexicanum (Hitche.) Beetle, Phytologia 54:
4. 1983 = Nassella mexicana (Hitche.) Pohl, Taxon
39: 611. 1990. [Basionym: — mexicana Hitche
Contr. U.S. Natl. Herb. 24: 247. 1925,]
NOMINA NUDA
Е avenaceus Raf. ех B. D. Jacks., Index Kew.
E 894, nom. inval., as syn. of Stipa avenacea E
ochaetium avenaceum (L.) Parodi.
Е dan Raf. ex B. D ). Jacks.,
ui 894, nom. vg .„„ as syn. of Stipa avenacea L.
iptochaetium avenaceum n (L .) Parodi.
Stipa sia Willd. ex Steud., Nomencl. Bot. (ed. 2) 2:
341, nom. inval., as syn. of Stipa avenacea
Stipa if Walter (1788) = Piptochaetium
avenaceum (L.) 1 li
Avena — Willd. ex Ste ii
146. 1841, nom. inval.,
"MH =
Hitche
— panicoides Trin., in Martius, Fl. id Enum. pl.
2(1): 376. 1829, Mém. Acad. ms Sci. St. Péters-
bourg s sér. а Sci , pro syn. nom.
nud. — Piptoc haetium montev idense (Spre ng.) Paro-
di.
Index Kew. 2
a rodi,
Bot. (ed. 2): 2:
Milium mexicanum
Piptoc hun fimbriatum (Kunth)
Nomencl.
syn. of
i Agrost.: 173. 1812,
Piptatherum elegans P. Beauv.,
ud. P — a DE E.
от. n
)
esv.
Ыш гит panicoides | Gilbert, Enum. Pl. Monte 17.
873, nom. nud. = Piptoc ает pu s а )
Е Desv.
ТУ — Gilbert, Enum. Plant. Montev.: 117.
73. nud. — Piptochaetium stipoides (Trin. &
n ) var. —
Literature Cited
Barkworth, 1986. — и Presl (Gramineae,
Stipeae) in — and Mesoameri Sans and dis-
tributional observations. Brenesia 25- 26: 169-178.
1988. New taxa in Piptoc — Sipe ae, Gra-
pa from Mexico. Syst. Bot. 13: 196-201.
——— 1990. Nassella (Gramineae, рене) Revised in-
terpretation and nomenclatural changes. Taxon 39:
7
. 1993. North American Stipeae (Gramineae): '
onomic c line 's and other comments. Phytologia 74:
25.
ax-
Beetle, A. А.
pns 37: pps
Brako, 1 ла 1993. Catalogue of the Flow-
ering — and Gymnosperms of Peru/Catálogo de las
eg siis y pn Yospermas del Perú. Monogr. Syst.
Missouri ua wd. 45.
Bremer, К. 198€ The limits of amino acid sequence data
in — rm phylogenetic reconstruction. Evolution
42:
— grasses from Mexico V.
--
EC Brane h support and tree stability. Cladis-
tics 10: 295-
Brown, №. V. ps ¢ hromosome numbers of some Texas
grasses. Bull. Bot. Club 78: 292-299,
Cialdella, A. M eee 1998. Revisión de las es-
pecies — ricanas del género Piptoc haetium (Poa-
ceae, Pooideae, Stipeae). Darwiniana 36: 107-15
Clayton, W. D. 1983. 212 Pipto-
Proposal to conserve
Volume 89, Number 3
2002
Cialdella & Giussani
Phylogenetic Relationships of Piptochaetium
chaetium К. B. Presl over Podopogon Rafi. (Grami-
‘axon 32: 649.
1853. ун In Ea
fisic ч у Ж a de Chile 6: 233—
Elias, M. K. 1932. Grasses = other — from the Ter-
tiary rocks of Kansas and Colorado. Univ. Kansas Sci.
Bull. 20: —
1942. ‘Tertiary Prairie Grasses and Other Herbs
from ‘he Bi Plains. Special Pap. Geol. Soc. Ame
M. Ne а Geological Survey, The University of Ne-
neae). 1
Desvaux, E. Gay (editor), Historia
. Lincoln.
Ellis. R. P 1979. A procedure for и compar-
alive leaf anatomy in the Poaceae. in epidermis as
seen in surface view. ое 12: Bl. 71.
Goloboff. Р. A. 1993
`2
puter prn a al [Distributed by the author. |
ONA version nit). Com-
93b. Estimating character weights during tree
search. gs — 9: 83-9]
1997. Pee-Wee version 2.8 (32 bit). жараш
— ме manual. [Distribute d y the author
Gould. )58. C hromosome numbers in lua "n
grasses. н J. Bot. 45: 757-768.
1966. С hromosome numbers of some Mexican
grasses. — Le ч 4: 1683-1006
Hite hcock. A. ‚ New species "d grasses from South
America. da roc. [^ Soc. Wash. 36: 195-198
. 19 — of the cal of ih Unite di States,
ed. in A. Chase. Misc. Publ. U.S. Dept. Agric.
—
Holmgren, P. K.. N. H. Holmgren & L. C. Barnett. 1990.
ndex Herbariorum. Part 1: The Herbaria of the World.
cara Veg. 120.
Jacobs W. L., J. Everett € M. Barkworth. 1995
C erui of morphological terms used in the Sti-
peae (Gramine * and a reassessment of Nassella ii
11.
Australia. Taxon 44: 33—
Jorgensen, P. M. & León-Yánez (editors). 1999. Cata-
logue of the Vascular Plants of Ecuador. Monogr. Syst.
Bot. Missouri Bot. Gard. 75.
Metcalfe. C. R. 1960. Anatomy of the Monocotyledons. I.
Gramineae. Clarendon Press, Oxfor
Munoz-Schick. M. 1990. Revisión del género Nassella
— К. De sv. (Gramineae) en Chile. Gayana. Bot. 47:
ue TS, E M. 1947. € — and genetics of forage grass-
s. 319-422,
Parodi, L. R. . Revisión de
americanas de fa género Piptochaetium. Revista Mus. La
10
c
las gramíneas australes
Plata, sece. Bot. 6: 213—
1961. Gramineas argentinas nuevas o críticas.
. Rev с Argent. Agron, 28(3—1): 100—
Репай е, . P. 1996. Anatherostipa. un nuevo género de
Poaceae. irene Bot. 53: 277-28
Presl, C. B. 18:
Prague I: x y ээ.
Reeder, J. R. 1968. Notes on Mexican grasses VIH. Mis-
cellaneous Chromosome numbers—2. Bull.
Club 95: 6 92 30.
1977. € ОЛЕГ numbers in western grasses.
Amien J. Bol. 04: M )2
Sánchez Vega, |. 19
del género Piptoc — J. 8.
[= V
). Reliquiae Haenkeanae. J. C. Calve.
Torrey Bot.
S visión de las espec ies peruanas
Presl (Gramineae).
—
i . Wasshausen & R. M. Klein. 1982.
Gê meros: 45. Deschampsia até 84. Pseude-
In R. Reitz (editor). Flora Hustrada Cata-
rinense | (Gram.): 443-906
Gramineas.
chinolaena.
Spegazzini, C. 1901. em Platenses. Anales Mus. Nac.
Montevideo 4: 1-173
Swallen, J. R. 1955. Flora af Guatemala 2: Grasses of
и . Fieldiana Bot. 24: 1-390
Thomasson. + R. 19
=
78a. E its К patterns of the lemma
in some fossil and living grasses and their phylogenetic
каа ance. Science 199: 975-977.
1978b. Te rs fossil plants — in Nebraska.
Res. Rep. Natl. Geogr. Soc. 1978 Projects: 553-501
280). — in — — (Gramineae
Stipeae): Paleophytogeograph Iselva
ignificance.
А ‚ Fossil grass anthoecia and other plant fos-
sils from arthropod burrows in the Miocene of western
)
silo adaptation
in ы live ла, ts (er mma Ар A аў. Апп. Missou-
ri Bot. Gard. 72: 843-851.
Torres, M. A. 1969. Nota sobre una i de Pipto-
alatum de Argentina у Uruguay. Bol. Soc. Argent.
Bot. 11: 251-254.
997. Nassella (Gramineae) del noroeste 1
— nlina, Stipa (Gramineae) del noroeste de la "ed ne
tina,
América del Sur.
Vicoraella ——
un nuevo género para
: Ministerio de la produc ción v el
empleo. provinc ia ie Buenos Aires, Comisión de ves-
tigaciones Científicas (С IC) — 13 16.
Valencia. J. 1. & M. Costas. 1968. E — citotaxonó-
micos wipe — him a ae ). Bol. Soc. Ar-
gent. Bot : 168-179
APPENDIX ]
Examined material of doe — Each specimen ts
cited by the last name of the f
more than one collector. Species number is indicated be-
st collector when there is
tween parentheses.
1. Piptoc ман alpinum 1. B. wa
2. Piptochaetium angolense Philip}
3. Piptoc adum angustifolium ( Mite heock) Valencia &
Costas
V. Piptochaetium avenaceum (L.) Parodi
5. Piptochaetium avenacioides (Nash) Valencia & Cos-
tas
Piptochaetium bicolor (Vahl) К. Desvaux
7. Piptochaetium se ы — —
8. Piptochaetium brevicalyx (E. Fournier) Ric
9. Piptochaetium burkartianum Parodi
10. Piptochaetium cabrerae Parodi
Piptochaetium calvescens Parodi
12. Piptochaetium confi i
13. Piptochaetium cuc ара Rosengurtt & Izaguirre de
Arturio
14. Piptochaetium featherstonet (Hitchcock) Tovar
15. ан (Kunth) Hitchcock
a nd hackelii (Arec havale 1а) Parodi
эз
т
4
N
N
=]
pes
"E
24. Piptochaetium napostaense (Spe
25. Piptoc haetium palustre Mujica-Sallés &
agner
». Piptochaetium panicoides (Lamarck) E. Desvaux
Longhi-
334
Annals of the
Missouri Botanical Garden
7. Piptoc haetium pilosum (Sánchez Vega) Cialdella &
Giussani
Piptoc haetium pringlei (Beal) Parodi
ruprechtianum E. Desvaux
N
28.
29.
30.
31.
32.
33.
33a.
Piptochaetium seleri (Pilger) Henrard
Piptochaetium setosum (Trinius) Arechavaleta
Piptochaetium stipoides (Trinius & Ruprecht) Hackel
Piptochaetium stipoides (Trinius & Ruprecht) Hackel
var. stipoides
Piptochaetium stipoides (Trinius & Ruprecht) Hackel
var. echinulatum Parodi
Piptochaetium tovarii Sánchez Vega
35. Piptochaetium uruguense Grisebach
36. Piptochaetium virescens (Kunth) Parodi
Aleman 267 (26); Alonso J. 151 (20); — ows s.n. (4):
Arriaga C. 351 (31); Arzivenco 349 (20), 352 (28).
5 (5), 1155 (5); Banda 132 (8); Barclay 8455
(34); Barkley E A 660 (15); Bartlett 10225 (33a); Bartley
2011 (4); Beaman 3928 (31), 4155 (31), 4273 (31), 4329
‚ 4387 (36); Beetle M-141 (36), M-4743 (36), M-7086
. M-7338 (15), M-7742 (15), M-7856 (28), М-7864
; Berkley 1579 (4); Biswell 94408 (4); Boechat 160
, 161 (20); Bonner 354 (4); Boott s.n. (4); Boufford
182: 12 (4); Bozeman 8842 (4); ipid — 31 (15), 34 (15);
Brant 2239 (15); Braun 1422 ( J42 (4): Breedlove
14171 (36), 39241 (36), 43293 a 44133 (15); Bro. Ar-
sene 2699 (36), 5305 (36), 5375 (36), 6749 (36); Brunken
2 м (15), 247 (36), 387 — gis ss 221 (4).
Calzada 4645 (15); Carter 2396 (15); Castillo ae
(31); Chapman 1037 (4); Chase 7220 (4), 7312 (4), 7
(4); Chazaro 3792 (31); С — 9077 (15); С DURAM 9 B
"ae Churchill 123 (4), s.n. (4); Cole s.n. (
.n. (4); Combo 876 (4); vens 2069 (4); Correll 353
(4), 1 3699 (15), 16170 (4); € ronquist 4311 (4), 4432 (4).
5014 (4), 5075 (4); Curtiss 5834 (5
Darrow 1287 (15); Davidse 9287 a 5), 9485 (15), 9832
(36), 10031 (15), 11137 A (33a), 29775 (15); De Chalmot
s.n. (4); De Koninck 73 (36), 114 (31); Deam 20709 (4),
4), 40704. (4), 64413 (4); Demaree 14586 (4).
19003 (4). 63483 (4); Dowell 7309 (5); Duncan 9485 (4);
31).
33b.
34.
=
Earle 165 (15); Early s.n. (4); Eggert 36 (4), s.n. (4);
Egler 41-136 (4); Ellis 1113 (31); Escobedo 1478 (31).
Faircloth 5079 (4), 5798 (4), 7070 (4); Felger 92-794
(15); Fernald 7760 (4); Fernández 734 (28), 735 (20); Fer-
ris 2736 (15), 2812 (15); Fisher 108 (31), s.n. (31). s.n.
(4); Fogg Jr. 2224 (4); Fredholm 5092 (5). 5725 (5); Free-
man 54117 (4).
Garber s.n. (4); Gentry 8492 (15); Gleason 8537 (4):
Godfrey y 3686 (4), 8085 (4), 67795 (4); Gould 2456 (28).
4419 (15), 9211 (8), 9602 (15). 9661 (15).
‚1 5409 (4); Griffiths 94 (15), 107 (28), 4836
, 7032 (15); Griscom 16409 (4); Guerrero
2 (36): € ;utte 185 (36)
p^ 164 (4); Harkins 809 (4); Harvey 1038 (3), 1458
(15), 1764 (4), 1854 (4); Havard 26 (15), s.n. (15); Hen-
derson 91-03 (4), 94-29 (4); Henrickson 15143 a (15),
15226 (15); Hermann 3201 (4), 4198 (4); Hernandez Xo-
locotzi X-2443 (15), X-2828 (36), X-2836 (36); Hill
14632 (15); Hinckley 245 (28), 493 (28); Hinton 1336
(36), 1893 (36), 2315 (36), 2316 (36); Hitchcock 520 (4),
812 (5), 813 (28), 837 (15), 967 (5). 1271 (4). 1318 (4).
1341 (4), 1346 (4), 1903 (28), 3715 (15), MES (15), 5947
(36), 5965 (36), 5973 (36), 6019 (36), 6266 (36), 6733
(8). 7151 (36), 7169 (36), 7677 (28), 7681 (28), 7741 (15),
12616 (4), 13511 (15), 13553 ү“ s.n. (4); Hoffmann s.n.
4); Hoge 260 (15); Hoogstraal 991 (36); House 2101 (4).
Iltis 213 (15), 214 (8), 958 О, 23032 (4); Innes 1130
D
(15).
Jacob 815 (4); Jones 479 (36), 24157 (4).
Kearney 10114 (15), 10541 (28); Kearney Jr. 1102 (4);
Kellogg 82 (4); King 3378 (31), S 32 (36), 5151 (36);
Kneucker 564 (4); Knieskern s.n. (4); Koch 762 (36),
71 10 (1: 5): Kral 2065 (4). 25586 (15), 34987
45934 (4), 46012 (4), 46479
1). 55030 (4); Krauc h 6495
Larrick 249 (4); Latham 2755 a. 4087 (4), 22391 (4);
Le Doux 2086 (31); Le Sueur 143 (15); Leavenworth 1077
36); Lemmon 2923 (15), 4678 (2 8); 1 — 11 м (26);
ecc pa 2121 (1), 2497 (1); Lundell 13153 (15);
Luz 39 (26), Lyonnet 48 (15), 62 a (36), 1826 (36). 1877
(36), 1948 (31), 2498 (36), 2511 (31), 2512 (31), 2527
(36), 2684 (36), 2719 (31).
Mackenzie 3109 (4); Madrigal A. (15); — — TO
M. 450 (36); Marsh Jr. 629 (28); — 1063 (31);
thews s.n. (4); Matuda 18907 (36), 19202 (31),
24 (36). 25753 (36). 25755 (36),
(36), 29473 (36), 29776 (3), 31875 (36); McCarthy
(4); McDougall Ws (4); Me-
Ge p 57 (15), · (15); MeVaugh 12268 (36); Mearns
2573 (28): Nu i (5); Mejia S. 218 (31); Mendoza
ses | 5); Metcalfe 746 (15); Metzler 2 (36); Mez 3164.
(15). 3166 (15); Mick 87 (8), 180 (31). о үе Y — f B
Miller Jr. s.n. (4); Mohr s.n. (4); Moo 72 (15), :
(15). 4164 (15); Moore Jr. 1233 (31). bus : 36). pene A en
Moorgen s.n. (29); Morong s.n. (4); Mueller 2342 (15),
2442 (15), 7948 a (15); Mujica Sallés 2 (20), 22 (28), 42
(26), 46 (26), 47 (26), 63 (33 a), 72 (26), 74 A (26), 81
(20), 83 (29), 130 (35). 141 (22). 142 (22), 144 (35), 151
(20). 155 (22), 166 (20).
Nealley s.n. (4); Neally 129 (15); Nee 22928 (36),
24775 (31): Nelson 1373 (36), 1373a (31): Noyes s.n. (4).
O'Neill 491 (4); Ortiz 1132 (36), 1194 (36).
Palmer 339 (15), 5154 (4), 5248 (4), 24541 (4), 31887
(15); Pedersen 5188 (35), 13333 (24); Peebles 3389 (15);
Pennell 17704 (15), 8929 (28), 18940 (15): Peterson 4010
—
T
25777
—
(15); Piedmont 4311 (4); Plank 53 (4); Pringle 18 (28),
502 (15), 639 (15), 3035 (15), 4759 (36), 5200 (31), 6236
(36). 6588 (36). 7606 (8), 8595 (15). 9574 (36). 13249
(31), s.n. (15); — 3589 (15).
Quarín 3163 (33a).
Redfield 4317 (4); Reeder 233 (4), 2221 (31), 3295
(15); Renvoize 3025 (20). 4143 (23); Reverchon 4128 (4).
s.n. (4); Rivero 1384 (23); Robinson 1 (4); Rodriguez 1284
(31); Roivainen 481 (11); Rolfe 43 (4); Rose 2750 (36),
5611 (86), Soon 8), 8700 (36), 8744 (15);
), 11940 (31). 16904 (15),
17547 5) 17704 15) : 35555 (31), 35863 (8), 50768
(8).
Sánchez Vega 2458 (34), 2769 (34), 2935 (30), 4984
(30); Sánchez-Ken 141 (15); Sandoval 68 (31); Schaack
2881 (4); Schinini 19229 (35); Schultz 41 (4), 14705 (23);
Scribner 138 (4), 266 (4); Scully 652 (4); Serrato Sánchez
s.n. (15); Seymour 109 (4), 1060 (4); Sharp 44144 (31),
s.n. (4); Shinners 7029 (4); Shreve 7707 И 5); Silveus 727
(15), 785 (15), 3470 (28); Simpson 1); Smith 926
(36), 3327 (18), 8302 (14), 10109 B (18), 10331 (26),
10488 (18), 10960 (34), 10964 (18), 11038 (18), 11794
A (18); 5now 6721 (31); Soderstrom 141 (28), 493 (36),
752 (15); Sohns 177 (36). 289 (15). 300 (15), 424 (15).
505 (15), 507 (15), 533 (15), 534 (15), 538 (36), 553 (8),
Volume 89, Number 3
2002
Cialdella & Giussan 3
Phylogenetic Helal bitine of Piptochaetium
581 (15). 601 (36). 603 (36). 604 (31). 782 (15). 790 (36),
799 (36). 804 (36). 819 (36). 977 (36). 982 (36). 1012
(31). 1016 (31). 1022 (15). An (15). 1147 (15). 1500
(36). 1535 (8); Solomon 1535 (4), 1710 (4). 11342 (23);
Sperry 370 (15); St. Pierre 888 (36); Standley 40512 (15).
10699 (15). s.n. (15); Stanford 2482 (4), 2604 (15): Ste-
vermark 3172 (15). 36700 (36). 48274 (31). 48452 (31).
50955 (36); Strong 1342 (4); Svenson 10246 (4); Swallen
951 (4). 1109 (15), 7561 (35).
Talbot 8 (4): Tavlor 276 (15).
(15): Tharp 4175 (28): —
28558 (4): — ye (4). ; Toolin 2222 (15):
Torres 875 (36), 1226 (29); e s * ). 4544 (4). 7031
(5). s.n. (4): s 61 r Tuckerman 848 ( ji
Valencia 442 (22), s.n. (23): Valls 1233 (20). 2
2805 (20). 2821 (33 a). An (20), 2939 (33 a),
(22). 12259 (35); Van Eseltime 127 (4): Van Schaack
(4): Van Sickle s.n. (4); Vasey s.n. (
(31); Ventura A. 376 (36). 2153 (15). 18506
5). 1921 (15). 1922 (28), 1959 Ов,
(15); Villarreal I1: (28).
Warner s.n. (4): Warnock 12082 (28), 46717 (15): War-
nock T 598 "e Pies: 6322 (15). 12627 (15). 14854
(4). 16410 (36): oo 1122 (4). 5831 (4). 6234 (4):
Weatherw rways E (4). 37 (4); West 26673 (4); W — 1
(15). 88 W "y ‘ox s.n. (15); Wolff 4974 (4): Wooton
$; E (15): ou 643
Yacolucei 843 ( та
Zanin s.n. (1). 18 (1); Zavala Ch.
Zöllner 8672 (32).
6254 (15): Tenorio 1197
үз ee (4). 22502 (4).
—
215 (20).
12188
3575
2 15 55
—
245 (31). 862 (15):
APPENDIN 2
cimens of Hesperostipa and Nassella used for mor-
phological and cladistic analyses, as outgroups of Pipto-
chaetium.
Hesperostipa comata (Trinius & Ruprecht) Barkworth
(= Stipa comata Trin. & Rupr.).
U.S.A. Colorado: El Paso Co., Ehlers 7526 (UTC
vada: Elko Co., E side of Ruby Mtns., 0.7 road mi.
Shantytown, Williams et al. 84—49—8 (UTC). Utah:
field Co.. located at the head of South Fork, Dugout Creek
on the Mes AM in the Henry Mountains. Velson
(UTC
Hesperostipa neomexicana (Thurber) Barkworth (=
Stipa و nnata L. var. neomexicana Thurber).
U.S.A. Arizona: without locality, dec 3051 (BAA).
Texas: San Antonio, Silveus 1215 (BA
1). Ne-
S of
Gar-
Hesperostipa spartea (Trinius) Barkworth (= Stipa spar-
te K — ).
S.A. Colorado: Bookvale. Clear Creek Co.. Chur-
PR s.n. (BAA 15155). Hlinois: Stark Co., Wady Petra,
Chase 808 (BAA). Indiana: — Co.. sandy soil along
Pennsylvania railroad and hwy. 29, 2.3 mi. S of Winamac.
Potzger 4243 (B:
Nassella charruana (Arechavaleta) Barkworth (= Stipa
charruana Arechav.).
ARGENTINA. Entre Rios: Gualeguaychú. Burkar
25865 (SI); Gualeguaychú, ruta 12, km 180, Burkart &
Crespo 22895 (SI).
Nassella leucotricha (Trinius & Ruprecht) R. W. Pohl
(= Stipa leucotricha Trin. & Rupr.).
XICO. С i ‚ Sierra de Arte: aga. cañón
E de Jame. — 249 (UTC). U.S.A. Ca
near Sonoma. Heller 5350 (Sl). Tex-
de Jame, :
ifornia: Sonoma Co..
Karnes Co.,
[sel
Texas 2154 (S1):
S of Karnes City,
as: Bee Co., Herb. Univ.
roadside wet ve immediately
10634 (UTC).
Nassella meyeniana (Trinius & Ruprecht) Parodi (=
Urachne meyeniana rin. & Rupr.).
ARGENTINA. Jujuy: Coe T а. Ahn a Pampa. Venturi
9385 (SD. Humahuaca: Tres Cruces, Parodi 9609 (S1):
Sierra del Aguila, Venturi € (SI).
Nassella pubiflora (Trinius is ape cht) E. Desvaux (—
Urachne pubiflora Trin. & Ruj
ARGENTINA. Jujuy: ipa CRM ‘a, camino a Pale а de
Aparzo, Kiesling & López 3624 (SI). Tucumán: Tafí, km
82. Quebrada del Barón, Diers 282 (S1).
Nassella pulehra (Hitchcock) Barkworth (= Stipa pulch-
ra Hitche.
S.A. California Marin Co., 5 mi. W of Mill Valley.
Beetle 2722 (SI): Contra Costa Co.. Orinda, in sandy soil
of open cut, Beetle 1718 (SI); Marin Co.. San Rafael Hills.
Howell 16251 (UTC
Nassella Tungane E.
CHILE. IV Región: — Yiles 472 (SI).
Nassella tenuissima (Trinius) Barkworth (= Stipa tenuis-
sima Trin.).
MEXICO. Veracruz: between Tziutlán and
Beetle M-581 (UTC). ARG ENTINA, Córdoba:
odoro, Stuckert 198 (SI). om Toay, Ea. A
Rúgolo 1060 (51). Tucumán: Tafí del Valle. Hueck 9 (S1).
Perote.
Nassella tric 'hotoma (Nees) Hackel ex Arechavaleta (=
Stipa АИ Nees).
ARGENTINA. "ui os Aires: Bolívar, Pirovano. Bur-
kart E (SD: Tandil. Troncoso 1290 (У). Entre Ríos:
E. Carbó. orillas de vías férreas. Burkart 18129 (S1).
APPENDIX 3
List of characters and character states.
the characters and character states and their range of var-
iation are presented as considered in the cladistic m
\ description of
l. Lemma margins: not involute = 0: involute =
2. Palea shape: flat on abaxial surface = 0: bikee : d on
the ab os surface =
3. Palea length: half the length of the lemma to equal
ше length of the lemma = 0: longer than the lemma
l: s E than half the le — of the le ‘mma = 2
(Fig. 2 2A.
1. Palea ve ^ nerveless = 0; 2-nerved = 1.
5. Long fundamental ce à of the lemma e Pm rmis: ab-
sent = Û; prese
6. Fundamental cel e walle irregularly sinuate = 0:
regularly sinuate = 2. Epi-
dermal characters (5, 6). were taken from Thomasson
(1978a). Barkworth (1990). and Torres (1997).
Callus length: shorter than 0.8 mm = 0; equal to or
longer than 0.8 mm = 1. Callus length is variable
within Piptochaetium; it is usually long in species
with cylindrical or obconical florets but short in ob-
ovoid or lens-shaped florets. In Nassella the callus is
variable in length, while it is always long in Hespe-
regularly dentate = l;
rostipa.
8. Callus shape: acute or — ute = 0 (Fig. 5
blunt or truncate = 1 (Fig. 5F-H). The callus. vias n
short, is generally trunc die or blunt; if long, it is acute
or subacute. Callus shape varies also within Nassella
s.l.. while in Hesperostipa it is always acute. Charac-
336
Annals
M — Garden
Du
On
^
~
О.
1.
wa
ters 7 and 8 were traditionally considered as diag-
nostic to identify Piptoc haetium sects. Podonoson and
Piptochaetium (Parodi, 1944; Sánchez Vega, —
Cialdella & Arriaga, 1998). Piptoc haetium sect.
— generally included species with a long md
‘ute or subacute callus, while Piptochaetium sect.
Ppt haetium presented species with a short and
blunt callus
Callus — e: glabrous or slightly hairy = 0;
densely hairy = 1. Most Piptoc — ы ies have
florets with a densely pubescent callu 5A-G).
It is usually glabrous or only alinir о іп
Piptochaetium brevicalyx (Fig. 5H), P. calvescens, P.
> hirtum, P. don p montevidense,
panicoides, and P. tovarii. In Nassella the «
glabrous or pubescent, with trichomes densely to
slightly disposed, while in Hesperostipa the callus is
always densely pubescent.
cucullatum,
P
allu us is
Awn: persistent = 0; deciduous = 1.
йс п e branches densely flowered = 0; not
densely flowered = 1. This character varies among
species of Piptochaetium = Nassella, in which the
branches may be denseh loosely diene’ while
in Hesperostipa branches are loosely flowerec
Upper glume le ngth: — than the floret = 0: equal
to the floret = 1 s of Piptochaetium, Nassella,
and ени" ban upper glumes longer than the
‘ly or
)ecie
floret. Only in P alpinum "id = Drei eal is the
upper — as ine as the floret.
Floret — rete to fusiform = 0 (Fig. 5А—С); ob-
conical = — & Arig, 1998, dh вв. п)
obovoid to eh bo = 2 (Fig. 5H): lens-shaped =
(Cialdella & Arriaga, 1998; fig. 10C, D). Floret —
varies among spec ies of Piptochaetium, being terete,
fusiform, obconical, obovoid, or lens-shaped. Nassella
presents terete, obovoid, or lens-shaped florets, while
in Hesperostipa florets are always terete
Lateral compression of the floret: — or slightly
compressed = 0 (Fig. 5A-H); conspic — com-
— = 1 (Cialde lla & Arriaga, 1998: figs. ip
10C, This character is variable among species of
Piptoc htm and Nassella, while in —— *
lorets are never laterally cor mpressec
Lemma pubesc ence: absen ) (Figs. ТА. 5H); pres-
ent = | (Figs. 1B, 5
Papillae of the lemma: then = Ө; present = 1 (Cial-
Arriaga, 199 . 2D. 10C). Papillae are
ew uan ies — — ^aetium and Nas
sella, but are alway ipa
Prickles of the lemma: absent
).
s absent in Hesper
= 0; present = | (Fig.
1
Prickles over b all the lemma surface: absent —
0; present = 2D).
Prickles on is 2s third. of the lemma: absent. —
. B). Hooks are similar
to pric — although these appendages differ slightly
(Ellis, 1979). These characters
mologous by Metcalfe (1960). Intermediate forms can
also be found, making it a difficult chari acter to c ‘las-
sify.
0; present = | (Figs. 2€, 5A
were rec ognized as ho-
some species of Piptochaetium, Nassella, and all spe-
cies of He sperostipa. Distribution of prickles (с пагас -
ters 18, 19) varies among species, although it is
ug
ways constant within species. Only the быш. of
2
22.
23.
N
=
20.
N
=J
— sin P — var. stipoides (Cialdella &
лара, 1998: fig. LOF-I) is highly variable, even
within a specimen.
Base of prickles: surrounded but not covered by adja-
cent epidermal ce a = 0; partially covered by —
epidermal cells = 1 (Cialdella & Arriaga, 1998:
3D). Partially covered prickle bases occur only in Pip.
toc — stipoides var. echinulatum Parodi.
Lemma width j the crown: as ed as the
crown = 0; narrower than the crown = 1. This char-
acter is related to floret shape: lemmas of cylindrical
florets are usually as wide as the crown or slightly nar-
rowed, while in obconical, obovoid, and lens-shaped
florets lemmas are conspicuously narrowed below the
crown, being more conspicuous in gibbous florets.
Crown contracted to 2 base of the awn: absent — 0;
present = 1 (Fig. The crown concept was dis-
— by diffe E ator (Spegazzini, 1901; : Parodi.
Вн, 390; Muñoz-Schick, 1990; Jacobs
el D 95; Cis dell & Arriaga, 1998). As inter-
preted м ге, it refers to the distal portion of the lem-
ma, which is fused to the base of E awn (Barkworth,
1990; Cialdella & Arriaga, 1998). In Piptochaetium
the crown can be contracted or when not con-
— it exceeds the — of the awn (Cialdella
Arriaga, 1998: figs. OF, . Nassella and Hes-
pones. present a contracte T crown, bei ing me mbra-
nous, — parted, and,
ing at tli base of es aw
wn bos — not revolute =
Arriaga, 1998: fig. ТОЕ); re — toward the outside
Cialdella & Arriaga, 1998: fig. ӨК). Some van
cies of Piptochaetium show a s e crown like a
ring. generally covered by pri əks, and some-
times macrohairs, In РАТА and Hesperostipa the
crown is always straight.
Macrohairs in p crown: absent = 0 (Fig. 2A); е
ent = | (Fig. 2C). This character is only variable
P. stipoides var. stipoides and P. fimbriatum. Thes 'se
appendages are frequently associated with other types
of epidermal appe ndages such as и ‘kles ч hooks.
Papillae in the crown: absent = 0; present = 1 (Cial-
della & Arriaga, 1998: fig. 3C). These penus are
found 1 in a few j W 8P ecies of — — while еу
ust. below
^
—
ickles, hoc
x. 2A); present
. This lype of Pere appendage
eral species of Piptoc Dan а an id
a few sy
is frequent in sev
Nassella, and complete sly — nt in
Hesperoaiina it is always pres nt.
/ ei e thread-like (E 5H); ja d
Aaldella & Arriaga, 1998: ig. ӘП). In Nassella,
di and almost all spec ies of Piptoc —
A awn is thread-like, bigeniculate, shortly hispid,
nd twisted in its basal portion. Only in Piptochae:
tium — is the awn cor me-shaped, 1 ight or
ес ie
="
=
slightly curved but not twisted, its base as wide as
* crown p ).9-1 mm wide).
Floret length/width ratio: > 3.6 = 0; < 3.5 =
Crown width: narrow [0,4—0 6(-0.9) mm diam. үз
wide (1-1.8 mm diam.) = 1.
crown is narrow when contracted
awn, except in teclea — which pre-
sents a wide and contracted crown (Cialdella & Arria-
ga, 1998: fig. ӨК) due to the conic ical she ape of the awn.
In Nassella — Hesperostipa the crown is always nar-
row in shap
SYSTEMATIC REVISION AND Silvia S. Denham, Fernando O. Zuloaga,
PHYLOGENY OF PASPALUM and Osvaldo Morrone?
SUBGENUS CERESIA
(POACEAE: PANICOIDEAE:
PANICEAE)!
ABSTRACT
y-five species are treated in this work, in which exomorphological characters are analyzed cladistic ally. Species
of ei subg. Ceresia are characterized by their rigid. filiform to lanceolate blades, inflorescences with one to
several racemes, rachis of the racemes winged and hy: dw to membranous, spikelets pilose, occasionally glabrous, with
the upper anthecium pale, hyaline to membranous, occasionally chartaceous, and the upper le — not enclosing the
tip of the upper palea. Species grow in South America from Mexico to Argentina and Uruguay. A cladistic — of
subgenus Ceresia was conducted to test its monophyly, and to establish its relationship with о vila groups of Paspalum.
A key to the species in subgenus Ceresia is given. as well as morphological description and illustration, and distribution
maps.
Key words: America, Ceresia, cladistics, Paspalum, Poaceae.
Paspalum L. includes approximately 330 species um subg. Ceresia, considering vegetative and
distributed in tropical and subtropical regions of productive characters, analyzing habitat and distri-
America, with a few taxa growing in the Old World bution range. A cladistic analysis was further per-
(Clayton € Renvoize, 1986). Due to its large num- formed. in order to test the monophyly of the
ber of species. and the morphological variation subgenus and to establish the relationships. of its
present within the genus, Paspalum has been di- species.
vided into subgenera. sections, or informal groups
mainly on the basis of morphological characters, HISTORY
such as inflorescence type or features of the spike-
let (Nees, 1829: Doell. 1877: Chase. 1927: Pilger,
1929; Clayton & Renvoize, 1986; Morrone et al
1995. 1996, 2000; Cialdella et al.. 1995).
Chase (1929) divided Paspalum into subgenera
Paspalum and Ceresia (Pers.) Rehb., characterizing
the
Ceresia was established at the generic level by
Persoon (1805—1807), on the basis of C. elegans (=
^ Paspalum membranaceum Lam., non Walt.. 1788
Subsequently, Reichenbach treated Ceresia as a
subgenus of Paspalum (Chase, 1929). Several spe-
cies were described in Ceresia in the 19th century
atter by its foliaceous rachis, with one to sev- (Flüggé, 1810; Trinius, 1826. 1828-1836. 1834:
eral racemes per inflorescence and spikelets dense- Nees. 1829; Doell, 1877)
ly pilose, with long white hairs, mainly along the
Nees (1829) considered, within Paspalum. sec-
margins of the upper glume and lower lemma. tion Digitaria, including P. pectinatum and species
Taxonomic studies i Ceresta are fragme ‘магу;
=
currently treated in Axonopus; section Lanigeri, in-
cluding P. eucomum, P. guttatum, P. erianthum, P?
sanguinolentum, P. ammodes, P. polyphyllum, and
P. blepharophorum (= P. polyphyllum); and sections
with respect to the subgenus, these Paspalum spe-
cies have been analyzed in regional works or floral
treatments (Nash, 1912: Chase 1927, 1929: Pilger,
1929: Hitchcock, 1951: Burkart. 1969: Rosengurtt— Cristati, Genuini. Axonopodes, and Ceresiae, with Р.
et al., 1970: Sendulsky € Burman, 1978, 1980: da pyramidalis (= P. repens) and P. stellatum.
Silva et al.. 1979: Pohl, 1980: Judziewiez. 1990: Doell (1877)
Renvoize, 1998: Rodríguez. 1998).
The present nen treatment revises Paspal-
grouped Paspalum into three sec-
lions: sect. Eremachyrion, sect. Emprosthion, and
sect. Opistion, including in section Opistion the fol-
! We thank Vladimiro Dudás for preparing the excellent — and the staff of Instituto de Botánica си
T — rn cially Alejandra Garbini. We also thank the Consejo Nacional de Investigaciones Científicas y Teen
(CONICE T for a research grant (PIP 4440) and the National Ge — Society for research grants #6042-97, 6698.
A АЕ 68
th Istituto ga Botanic a Darwinion. Labardén 200, Casilla de Correo 22, San Isidro B1642HYD. Argentina.
sdenham larwin.edu.a
=
ANN. MISSOURI Bor. GARD. 89: 337-399. 2002.
338
Annals of the
Missouri Botanical Garden
lowing species: Paspalum aspidiotes (subg. Ceresia),
P. pectinatum (subg. Ceresia), P. eucomum (subg.
Ceresia), P. blepharophorum (= P. humboldtianum,
subg. Ceresia), P. ammodes (group Eriantha), P. gut-
tatum (group Eriantha), P. sanguinolentum (group
Eriantha), P. erianthum (group Eriantha), P. stella-
tum (subg. Ceresia), P. membranaceum (= P. cere-
sia, subg. Ceresia), P. heterotrichon (subg. Ceresia).
P. trachycoleon (subg. Ceresia), P. lanciflorum
(subg. Ceresia), and P. carinatum (subg. Ceresia).
Bentham (1881) considered three sections in
Paspalum: sect. Anastrophus, sect. Cabrera, and
sect. Eupaspalum, the first two characterized by the
presence of distichous spikelets with the lower lem-
ma facing the axis, while in section Eupaspalum
spikelets are unilateral with the lower lemma out-
ward from the axis of the raceme. Within section
Eupaspalum, Bentham (1881) distinguished four
subsections: subsect. Anachyris, as an artificial
group with a single bract below the upper anthe-
cium; subsect. Ophistion, with species without a fo-
liaceous rachis; subsect. Seudoceresia, with a rachis
more or less foliaceous, green, and concave with
glabrous and small spikelets (including such spe-
cies as Р, stoloniferum and Р. repens); and subsect.
Ceresia, characterized by its foliaceous rachis, with
the margins membranous and colored, spikelets cil-
iate and larger. Within subsection Ceresia, Bentham
included many tropical species and also Р cymbi-
forme.
Pilger (1929) divided Paspalum into eight sec-
tions: sect. Eupaspalum, sect. Anachyris, sect. Pter-
olepidium, sect. Erianthum, sect. Cymatochloa,
sect. Ceresia, sect. Eriolepidium, and sect. Moen-
This author treated in section Ceresia such
species as P. membranaceum (= Р. ceresta), P. stel-
latum, P. heterotrichon, P. carinatum, P. eucomum,
P. pectinatum, Р. cordatum, P. humboldtianum, and
P. trachycoleon.
Chase (1929).
species of Paspalum, related subgenus Ceresia with
in her study of North American
her group Dissecta, both with the rachis of the ra-
cemes foliaceous and winged, separating Dissecta
by the presence of glabrous — Among the
species included by Chase in her treatment are:
Paspalum pectinatum, P. contractum (= P. lanciflo-
rum). P. stellatum, P. heterotrichon, P. trachycoleon,
P. cymbiforme, and P.
(1929) stressed that P sanguinolentum (excluded
humboldtianum. Chase
here) did not belong to subgenus Ceresia but to
group Eriantha. Later, Chase (ined.) plac ‘ed in sub-
genus Ceresia P. carinatum, P. ceresia, P. heterotri-
chon, P. stellatum, P. trachycoleon, P. phyllorhachis,
P. soboliferum (= P. humboldtianum), P. polyphyl-
lum, P.
humboldtianum, P. humboldtianum var.
P. humboldtianum), P. buchtienii, P.
2
stuckertii (=
malmeanum, P. eucomum, P. splendens (= Р. eu-
comum), P. guttatum, and P. ammodes. On the other
hand, Paspalum aspidiotes, P. pectinatum, P. cor-
datum, P. setiglume, P. contractum, and P. lanciflo-
rum were grouped in section Pectinata (see Table
4 for groups considered here).
Clayton and Renvoize (1986) recognized eight
sections within Paspalum: sect. Diplostachys, sect.
Pterolepidium, sect. Anachyris, sect. Erianthum,
sect. Eriolepidium, and sect. Moenchia with a tri-
quetrous, not membranous, rachis: sect. Paspalum
and sect. Ceresia with a membranous or foliaceous,
winged rachis. Section Ceresia was distinguished
Clayton & Renvoize, 1986) by including species
—
with a membranous, colorful rachis, and spikelets
with the upper glume and lower lemma scabrous,
usually ciliate, sometimes winged.
Rodríguez (1992) established Paspalum sect.
Pectinata and distinguished it from section Ceresia
by including species with spikelets winged and
broadly ovoid, cordiform or not at the base
(1994) dese ире, within
Paspalum subg. Ceresia, section Biaristata, includ-
Filgueiras and Davidse
ing two species: P. biaristatum and P. longiarista-
tum, both defined by conspicuous awns on the up-
per glume and lower lemma.
Most recently, Rodríguez (1998) made an ac-
count of nine species of Paspalum subg. Ceresia for
Venezuela: P. setiglume, P. pectinatum, P. lanciflo-
rum, P. carinatum, P. humboldtianum, P. polyphyl-
lum, P. stellatum, P. trachycoleon, and P. heterotri-
chon, all with a winged rachis and pilose spikelets.
MATERIALS AND METHODS
MORPHOLOGICAL ANALYSIS
Specimens were examined from the following
herbaria: B. BAA, BM, COL, CONC, CTES, G,
— LIL, LPB, MA, MEXU, MO, P, R,
‚ US, W (Holmgren et al., 1990; Ap-
pendix PS | list of specimens examined for the
outgroup taxa is provided in Appendix 3. Scanning
electron micrographs (SEMs) were prepared of the
upper anthecium of six species, utilizing the pro-
cedures described by Soderstrom and Zuloaga
1989). The
marked with an asterisk in Appendix 2: the type
specimen of P. reticulinerve, Solomon 17003, was
vouchers for this SEM study are
—
also used. The specimens were viewed on a Zeiss
940 A scanning electron microscope of the Dar-
winion institute, operating at 10—20 kV.
PHYLOGENETIC ANALYSIS
Characters. Fifty-eight exomorphological char-
acters were used in the cladistic analysis: 15 char-
Volume 89, Number 3
2002
Denham et al. 339
Paspalum subg. Ceresia
Table 1. Morphological characters and character states
used for the cladistic analysis of Paspalum subg. Ceresia.
Life form
1. Life cycle: perennial (0), annual (1). This character is
a sy napomorphy for — Racemosa in Paspalum.
es in section — are Paspalum lon-
uum ang P. cac —*
uatic or — aquatic plants (1).
ıs with spongy Sith: absent (0), present (1).
Both characters 2 and 3 are helpful to solve relationships
within the outgroup.
Blades
4. Shape: linear-lanceolate (0), linear (1).
э. Basal leaves: not. pseudopetiolate (О), shortly pseu-
dopetiolate (1), markedly pseudopetiolate (2). Paspal-
um buchtienti, P. cymbiforme, anc
have s
Annual specie
ві aristi
2 —
shortly pseudopetiolate leaves, with the pseu-
dopetiole a narrowed zone at the base of the blade.
Markedly pseudopetiolate leaves is an autapomorphy
a long narrowed
zone between the — and the flattene d blade: both
about the same leng
for P. imbricatum, where leaves
>
Inflorescenses
First-order branching of species here treated in Par
cum and Anthaenantiopsis are considered аа to
unilateral racemes in Paspalum.
6. Spikelets b arranged on the rachis: absent
(0). present (1). l inilaterally arranged spikelets are
present in all studied species alum.
Truncate inflorescences: oni (О), present (1). The
inflorescence is truncate whe i
end in a spikelet, and the most distal raceme is a first
order branch
8. Raceme number per inflorescence: 1 (0). 2 (1), 3 or
more racemes (2).
9. ннан, of racemes: solitary (0), conjugate (1),
subdigitate (2), alternate (3). Racemes are conjugate
1 the internode between two contiguous racemes
is almost absent; when these internodes are short, up
lo finally, when
the internodes are well developed, racemes are con-
ст long. racemes are subdigitate:
sidered as alternate
10. Rachis of the racemes: triquetrous (0), winged with
longitudinal veins at — wings (1), winged with anas-
tomosed veins at the wings (2), winged and nerveless,
with hyaline to кр Мокен wings (3)
scored as triquetrous when the ventral side has the
e rachis was
same development of the lateral margins. while it was
s winged when the lateral margins are
Nerveless and haine to
m ашин wings i» occurs in some species of
considered
clearly developed and flat.
‘the racemes: duni ina spike let (0), in a
duis du (1). in both a apike ‘let or in a naked point
let or in a naked boit in several species of * sub-
genus, while in P. goyasense it always ends in a
developed spikelet
12. Rachis disarticulating at maturity: absent (О), present
(1). This character is a synapomorphy of the Race-
mosa group.
Table 1. Continued.
13. Pedicels laterally inserted in the spikelets: absent (О),
present (1). This — i
erotrichon, P. petr rachycoleon, and P. phyl-
lorhachis, where ar is — at the lateral
base of the upper ane (Fig. ТЕ,
Spikelets
14. Shape: biconvex (0), plano-convex (1). Plano-convex
sp vikelets are a EM within the studied spe-
cies of Paspalur
15. Disposition — (O), paired (1),
lower spikelet reduc ed or r aborted
paired with the
2); with 3 or more
ГС
le are present on one а h in п Pi
icum = and P. obtusum.
16. Spikelet with an annular thic kening at the — ab-
sent (0). p esent (1). This character is present in Pas-
palum — thon, P. petrense, P. — and
P. phyllorhachis.
Lower Glume
17. Presence: present (0), absent (1). The lack of a lower
glume is a synapomorphy of the studied. species of
p aspalum
Upper glume
18. Texture: hyaline (0), ии We consider Ше
de ght passes through it.
present. (1).
winged upper glume ST flat and extending margins.
This character is shared by n Pectin-
ata, with — exception és Paspalum paper Md and
? cachimboe
20. Base — als n (0), prese nt (1). As in the pre-
vious character, this feature is present in section Pec-
tinata (but mot: in Ты — нун * P. cach-
imboense).
species of sectio
—
N
Apex: nol awned (0). awned (1). Awned glumes are
present in Paspalum longiaristatum and P. biarista-
tum.
22. Corky margins: absent (0), present (1): usually found
in Pospalum buchtienti, P. суть P. humbold-
tianum, P. polyphyllum, P. niquelandiae. and P. bur
manii.
iforme,
). We consider
when they
23. Ciliate margins: absent (0), present (1
that the margins are ciliate
rigid, and setose hairs in only one series, Paspalum
reticulinerve and Р. aspidiotes.
24. Pilose margins: absent (0), present (1). Mu eee I
have short
> long, not rigid, and arranged in s èveral s
25. Radiate marginal hairs: absent (0), present m A hen
marginal hairs are radiate, they look like a crown, as
in Pom buchtienii, p homo. P. poly-
phyllum, P. аша а and P. burmanit.
26. Surface pilosity: absent (0), hairs — arising from
і basal point (1), bus present in the lower half of
the жые glume (2), hairs all over the surface of th
upper EM (3). Hairs are usually found in Paspalum
78
g. Ceresia in the basal portion and margins of the
übt glume; they are less frequent in the distal por-
ion.
Annals of the
Missouri Botanical Garden
Continued.
Table 1
Table 1. Continued.
2 (0), 3 (1), 5 (2), 7 (3).
equidistant (0), lateral nerves
near the margins (1). This character is helpful to re-
solve relationships within outgroups.
29. Position of lateral nerves: marginal (0), submarginal
(1).
30. Lateral extension of the margins: flat (0), curved after
the nerve (1), plicate along the nerve (2), absent (3).
We considered a lateral extension present when the
margin continues after the lateral nerve.
31. Marginal nerves only present in the lower half of the
. This is a com-
s on Pectinata, with
the —— of Passo bn iflorum and P. cach-
27. Numbe в of nerves:
=
=
-
=
=
к
ч
ys
=
=
=
=
=
-
=
=
> ы
У.
А
=
—
—
<~
=>
w
~
4
ج
2
элш;
—
imboens
32. id veins in the upper glume: absent (0),
present (1). This is a character observed in Paspalum
reticulinerve and P. aspidiotes.
33. Base ко шар, absent (0), present (1). '
is observed in Paspalum buchtienii and P. а.
tianum, both species with basal margins extended.
Lower floret
34. Lower floret: present, staminate (0), absent (1). The
absence of a lower floret is a synapomorphy of Pas-
patum
Lower lemma
35. Texture: hyaline (О), membranous (1).
36. Wings: absent (0), present (1). Wings on the lower
lemma were considered present when the margin of
the lower lemma is flat and extended, as in rene
reticulinerve, P. imbricatum, and P. aspidiot
37. Base: rounded (О), subcordate (1), cordate (2). Ad
date lower lemma was observed in Paspalum imbri-
catum and P. aspidiotes, while it is subcordate in P.
pectinatum and P. cordatum.
38. Apex: not awned (0), awned (1). awned apex o
the lower lemma is found in ae longiaristatum
and P biaristatum.
39. Ciliate margins: absent (0), present (1). As in char-
acter 23, the lower lemma margins are considered
ciliate when these appendages are short, rigid, setose,
and arranged in one series along the margins, as in
Paspalum reticulinerve and P. aspidiotes.
40. Margin pilosity: absent (0), present (1).
41. Tuberculate, marginal hairs: absent (0), present (1).
This is a character observed in / ie lane —
P. cachimboense, P. pectinatum, and P. cordat
42. Surface pilosity: Absent (0). ent
43. Number of nerves: 2 (0), : . 5 (2).
14. Disposition. of the nerves: — nt (0), lateral
nerves near the margins (1). This character is helpful
to resolve relationships within outgroup
45. Disposition of lateral nerves: marginal (0). submargin-
al (1).
46. Lateral extension of the margins: flat (0), curved after
the nerve (1), plicate along the marginal nerve (2),
absent (3). See character 30.
47. Marginal nerves only present in the lower half of the
One or two mar-
ginal nerves, › not reach the apex, are present
in Paspalum imbricatum and P. aspidiotes.
fob
-
-U
۹
48. Relative length of upper glume and lower lemma:
both bracts of the same length (0), upper glume longer
1), lower lemma longer (2).
Upper anthecium
49. — — to lanceolate (О), i 0
Stipe: abse 0), present (1).
p aspalum ——— P. cac мй р aspidiotes,
and P. stellatum.
— i
ne apex: not gaping, when the
upper lemma encloses the apex of the palea (0), g
ing, when the apex of the palea is free (1). All species
of Paspalum subg. Ceresia have an open upper an-
thecium.
53. Relative length of the anthecium and the spikelet: equal
(0), upper anthecium shorter than the spikelet
54. Papillae at the upper lemma: absent (0), sin ple a
pillae (1), verrucose papillae (2). A simple papilla i
an epidermic prominence including the cellular lu-
men, and a verrucose papilla e a papisa with addi-
tional prominences on the surface (Arriaga, 1987).
Verrucose papillae are found in — guttatum,
in the Eriantha outgroup, and in some specimens of
P. stellatum within Paspalum subg. Ceresia.
55. * ‘a bodies at the upper lemma: absent (0), present
56. Unicellular macrohairs at the upper lemma: absent
(0). present (1).
57. Bicellular microhairs at the upper lemma: absent (О),
present
58. Lodicules: present ( (0), absent (1). Lodicules are ab-
sent in
biaristatum.
Paspalum ceresia, P. longiaristatum, and P.
acters are multistate and were treated as unordered;
11 of these characters were uninformative since
they did not show variation within the ingroup. Ta-
ble 1 provides the list of characters used in the
analysis. Autapomorphies were excluded from the
analysis. The data matrix is shown in Table 2. Some
potentially informative characters were left out of
the analysis, either because of the large amount of
intraspecific variation, the lack of enough material,
or because it was extremely difficult to score them
due to overlapping states in different species, e.g
spikelet length and width, rachis width, and blade
pilosity.
Cladistic analysis.
based on maximum parsimony using NONA version
2.0 (Goloboff, 1997a) under equal weights. Due to
he large number of terminals, a heuristic analysis
The cladistic analysis was
was run following command sequence: “hold 10000;
rseed 0; hold/40; poly = mult* LOO; max*;
sswap*." One hundred subsearches were performed,
each constructing a Wagner tree using a random ad-
dition sequence. swapping the initial tree with TBR
> amb-:
Volume 89, Number 3 Denham et al. 341
2002 Paspalum subg. Ceresia
Figure 1. 'anning electron mic uerographs of the upper anthecium of Paspalum.—A. P. — apex of the upper
lemma (Dusén ge SI). —B. P. reticulinerre, apex oi M = per palea (Solomon 17003, MO). —C. P. trachycoleon,
upper palea with simple papillae — 2046, US P. biaristatum, upper palea "D bicellular mic rohi vrs,
macrohairs (Oliveira 750, SI. E. KV. P. trachycoleon. 3 pus let base. dorsal view (Tamayo 2046, US). —F. S pikelet
base. ventral view (А — 3004. Pa
(tree. bisection-reconnection) and retaining a maxi- swapping cutting two sister nodes simultaneously.
mum of 40 trees in each replication. The resulting The default option “amb-” retained only nodes that
trees were swapped using one round of TBR (max*
are supported by unambiguous optimizations, while
sswap*.” which performed TBR poly = treats trees as polytomous.
we
and one round of
342 Annals of the
Missouri Botanical Garden
Table 2. Data matrix used in the cladistic analysis. Polymorphic character states are scored as: A = [01]; B =
[02]; € = [03]; D = [12]; E = [23]; F = [012]; G = [013]; Н = [023]; 1 = [123]. Dashes indic ‘ate character inapplicable
for a taxon.
1991). First, command “jump*3” was used, which
performs TBR swapping on trees up to 3 steps lon-
ger than initial trees (Goloboff, 1993). The second
strategy was carried out using “hold/5; mult*400,7
which performed 400 independent searches begin-
ning from different Wagner trees.
Bremer support (Bremer, 1994) and jackknife
Lor gd 1 1 1 1 |) 1 |
Character | 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
Panicum laxum 0 0 0 0 0 1 0 2 3 0 0 0 0 0 G 0 0 1 O0
Panicum obtusum 0 0 0 0 0 0 0 2 3 0 0 0 0 0 G 0 0 I O0
Anthaenantiopsis fiebrigii 0 0 0 0 0 1 0 2 3 0 0 0 0 0 I 0 о | 0
Paspalum acuminatum 0 | 1 0 0 |! | F H 1 0 0 0 | 0 0 1 0 O0
P. almum 0 0 0 0 0 1 l D 1 1 0 0 0 l 0 0 l 1 0
P. candidum I1 0 0 0 0 1| 1 2 3 2 1 1 0 1 0 0 ıl — —
P. cromyorrhizon 0 0 0 1 0 1-1 D 2 0 0 0.0 1.0 0 1 1 0
P. dissectum 0 l 1 0 0 l 1 2 3 2 0 0 0 1 0 0 1 A 0
P. erianthum 0 0 0 0 0 1| 1 2 E 1 0 0 O 1 1 ото O0
P. guttatum 0 0 0 1 O 1.1 F F 1 0 0 0 1 0 0 1 1 O0
P. notatum 0 0 0 0 0 l 1 D D 0 0 0 0 1I 0 0 | 1 0
P. prostatum 1 0 0 0 0 1 1 2 E 2 1 1 0 1 0 O0 | 1 0
P. racemosum 1 0 0 0 0 1 1 2 E 2 1 1 0 1 0 0 1 | 0
P. sanguinolentum 0 0 0 0 0 1 !|! D 3 1 0 0 0 1 2 0 1 I 0
P. aspidiotes 0 0 0 0 0 1 1 2 331 0 0 1 0 O 1 | l
P. biaristatum о 0 0 1 0 | | F C 3 1 0 0 1 0 0 1I 0 O0
P. buchtienii о 0 0 0 1 l | 2 E 1 2 0 0 1 2 0 1 A O0
P. burmanii 0 0 0 0 0 1 1 2 3 2 2 0 0 1 1 0 | 1 0
Р. cachimboense | 0 0 0 0 |! | F 2 3 1 0 0 1 0 0 1 ] 0
P. carinatum 0 0 0 1 0 | 1 A € 3 1 0 0 1 0 0 ! ] 0
P. ceresia 0 0 0 A 0 1 1 F € 3 l 0 0 l 0 0 ] A 0
P. cordatum 0 0 0 1 0 | 1 2 2 1 1 0 отоо 1 1
Р. cymbiforme 0 0 0 0 1 1 IFC 3 1 0 0 1.) 2 0 1 A 0
Р. eucomum 0 0 0 1 0 1 1 D D 3 1 0 0 1 0 0 ! ] 0
P. Ke 0 0 0 0 0 1 1 F € 3 0 0 0 1 0 0 1 1 0
P. heterotrichon 0 0 0 0 0 1 l F € 3 l 0 1 1 0 1 1 0 0
P. ern о 0 0 0 1 1 1 D 3 1 2 0 0 1 2 0 1 0 O0
P. imbricatum 0 0 0 0 2 | 1 D E 3 1 0 0 1 0 O 1| 1 1
Р. lanciflorum 0 0 0 0 0 | | F B 3 1 0 0 1I 0 0 | | 0
P. longiaristatum | 0 0 1 0 1 1I F C.3 1 0 0.1 0 0 1 0 0
P. malmeanum 0 0 0 ] 0 | 1 A A 3 1 0 0 1.0 0 1 | 0
P. niquelandiae 0 0 0 0 0 1 1 2 3 1 2 0 0 | 1 0 1 1 0
P. pectinatum 0 0 0 0 0 1 1 D D 3 1 0 0 1] 0 0 1 A |
P. petrense 0 0 0 0 0 1| г 2 S 2 1 0 1 | 2 ] 1 A 0
P. phyllorhachis 0 0 0 0 0 1 71 2 3 2 | 0 1 111 1 A 0
P. polyphyllum 0 0 0 A 0 1 1 F € 2 2 0 0 1 2 0 1 0 0
Р. reticulinerve 0 0 0 0 0 1 ID D 1 1 0 0 1 1 0 |! l
P. stellatum 0 0 0 1 0 ! 1 A A 3 1 0 0 1 0 0 ! 1 0
P. trachycoleon 0 0 0 0 0 | 1 2 3 2 1 0 1 l l l | A 0
Tests to evaluate homoplasy were consistency in- group frequencies (Farris et al., 1996) were con-
dex (CI) and retention index (RI) (Farris, 1989). ducted in order to calculate branch support with
Two strategies were carried out in an attempt to NONA. Commands used for Bremer support were
find possible islands for MP trees (Maddison, — "subN: hold1500; find*”; bsupport performing TBR
swapping on preexisting trees, saving trees up to * 4
steps longer at each round. Jackknife frequency
values were calculated using the available instruc-
tion file for NONA, the JAK.RUN file. One thou-
sand iterations were performed, randomly deleting
36% of the characters, with the search options
“mult*5,” and the FQ.EXE file was used to cal-
Volume 89, Number 3
2002
Denham et al. 343
Paspalum subg. Ceresia
Table 2. Extended.
2 2 2 2 2 2 2 2 2 2 j j 3 3 3 j 3 j j 3
0 1 2 3 1 5 6 7 8 9 0 1 2 3 4 5 6 T 8 9
0 0 0 0 0 0 0 D 0 l 1 0 0 0 0 1 0 0 0 0
0 0 0 0 0 0 0 3 0 l 0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 3 2 0 1 0 0 0 0 0 l 0 0 0 0
0 0 0 0 0 0 0 2 l l 0 0 0 0 1 0 0 0 0 0
0 0 0 0 0 0 0 2 1 1 2 О О 0 l l 0 0 0 0
— — — — — — — — — — — — — — ] 0 0 0 0 0
0 0 0 0 0 0 0 2 l | 2 О 0 0 ] | 0 0 0 0
0 0 0 0 0 0 0 2 l | 1 0 0 0 1 1 0 0 0 0
0 0 0 0 1 0 j 2 l l l 0 0 0 1 0 0 0 0 0
0 0 0 0 1 0 j 2 | 1 1 0 0 0 l l 0 0 0 0
0 0 0 0 0 0 0 2 l l 2 0 0 0 1 l 0 0 0 0
0 0 0 0 0 0 0 1 0 | l 0 0 0 1 l 0 0 0 0
0 0 0 0 0 0 0 1 0 | l 0 0 0 1 | 0 0 0 0
0 0 0 0 ] 0 3 2 | l 1 0 0 0 1 1 0 0 0 0
l 0 0 l 0 0 0 3 0 l 0 l l О 1 1 l 2 0 l
0 l 0 0 1 0 1 1 0 0 3 0 0 0 1 1 0 0 | 0
0 0 l 0 l | 2 | 0 1 2 0 0 l 1 1 0 0 0 0
0 0 1 0 1 1 2 1 0 l 2 0 0 0 l l 0 0 0 0
0 0 0 0 0 0 l l 0 | | 0 0 0 1 1 0 0 0 0
0 0 0 0 l 0 2 ] 0 | 2 0 0 0 1 | 0 0 0 0
0 0 0 0 1 0 l 1 0 0 3 0 0 0 1 0 0 0 0 0
l 0 0 0 0 0 0 2 0 | 0 1 0 0 1 1 0 l 0 0
0 0 | 0 l 0 2 l 0 | 2 0 0 0 1 1 0 0 0 0
0 0 0 0 l 0 2 l 0 l 2 0 0 0 l l 0 0 0 0
0 0 0 0 l 0 2 | 0 | 2 0 0 0 1 l 0 0 0 0
0 0 0 0 1 0 l 1 0 | 2 0 0 0 1 | 0 0 0 0
0 0 l 0 1 1 3 1 0 | 2 0 0 1 1 1 0 0 0 0
l О О 0 0 0 0 2 0 l 0 1 0 0 1 1 1 2 0 0
0 0 0 0 0 0 | 1 0 1 l 0 0 0 1 1 0 0 0 0
0 l 0 0 l О l l 0 0 3 0 0 0 ] 1 0 0 l 0
0 0 0 0 l 0 2 l 0 | 2 0 0 0 1 1 0 0 0 0
0 0 1 0 1 1 0 l 0 | 2 0 0 0 ] 1 0 0 0 0
0 ) 0 0 0 0 D 0 l 0 l 0 0 1 l 0 1 0 0
0 0 0 0 l 0 2 l 0 | 2 0 0 0 1 l 0 0 0 0
0 0 0 0 0 0 0 1 0 | 2 0 0 0 l 1 0 0 0 0
0 0 1 0 l | 3 1 0 | 2 0 0 0 1 1 0 0 0 0
1 0 0 l 0 0 0 j 0 | 0 l | 0 1 l l 0 0 l
0 0 0 0 1 0 1 \ 0 0 3 0 0 0 l l 0 0 0 0
0 0 0 0 1 0 2 l 0 | 2 0 0 0 l l 0 0 0 0
culate the majority rule consensus trees of the out-
put (Goloboff. 1993).
Separate analysis using implied weights (Golo-
boff, 1993) was run in PeeWee version 3.0 (Golo-
boff, 1997b) using the same search strategies as in
NONA. PeeWee weights characters according to
how well they fit a specific tree. Homoplastic char-
acters were downweighted in proportion to their
number of extra steps. We used a medium concave
function, K = 3.
In the present. treatment we fol-
Outgroups.
lowed Nixon and Carpenter (1993) for selecting the
outgroups. Species of Paspalum that share potential
synapomorphies with those of subgenus Ceresia
were included as outgroups: taxa of the Dissecta.
Eriantha, Racemosa, and Моа groups in Paspal-
um were chosen because they share with subgenus
Ceresia a wide rachis in the inflorescences (Chase.
1929, ined.: Morrone et al., 1995; Zuloaga & Mor-
rone. in prep.). Therefore, P acuminatum and P.
dissectum, of group Dissecta, P erianthum, P. gut-
tatum, and P. sanguinolentum of group Eriantha, Р,
Annals of the
Missouri Botanical Garden
Table 2. Continued.
4 4 4 4 4 4 1 4 4 5 5 5 5 5 5 5 5 5
Character 0 1 2 3 4 5 7 8 9 0 1 2 3 4 5 6 7 8
Panicum laxum 0 0 0 ро 1 0 0 0 0 D 0 0 0 O0 1 0 O0
Panicum obtusum 0 0 0 2 O0 l 0 2 0 0 2 0 0 0 1 0 1 0
Anthaenantiopsis fiebrigti 1 0 1 2 0 | 0 2 0 0 1 1 O 1 ! 1 | O0
Paspalum acuminatum 0 0 0 2 I| 1 0 0 0 0 2 0 1 1 1 1 1 O0
P. almum 0 0 0 2 | l ооо 0 2 0 l и 0 0 0 0
P. candidum о 0 0 1 0 1 0 — 0 0 D 0 0 1 1 0 0 0
P. cromyorrhizon 0 0 0 2 l l о 0 0 0 D 0 1 1 1 01 0
P. dissectum 0 0 0 2 l 1 0 0 l 0 2 0 0 l 1 0 0 0
P. erianthum 1 0 | 1 01 0 0 0 0 2 0 1 1 0 0 1.0
P. guttatum lO ! ]! 0 | 0 0 0 0 2 0 1 2 1 0 1 O0
P. notatum 0 0 0 2 | l 0 0 0 0 2 0 1 1 0 0 1.0
P. prostatum 0 0 0 1I 0 1 0 0 0 0 2 0 0 1 0 0 то
P. racemosum 0 0 0 1 0 |! 0 0 0 0 2 0 1 O т 0 0 O0
P. sanguinolentum 1 0 1 2 | I 0 0 0 0 2 0 0 I т O I O0
P. aspidiotes 0 0 0 2 0 1 1 1 O 1 |! 1 1 1 1 I |! O0
P. biaristatum 1 0 | 1 0 0 0 1 0 0 ! l 1 0 0 1 1 1
P. buchtienii 0 0 0 | 0 | 0 0 0 0 1 1 1 1 1 l | 0
P. burmanii 0 0 0 1 0 | 0 0 0 0 1 1 0 1 1 1 | 0
P. cachimboense | 1 0 | 0 1 0 0 0 | | 1 1 1 0 1 | 0
P. carinatum | 0 1 | 0 l 0 1 0 0 1 1 1 1 0 1 | 0
P. ceresia | 0 1 0 0 l 0 0 0 0 0 l 0 0 1 1 l 1
P. cordatum | 1 l 1 0 |! 0 1 0 0 ! 1 l 1 l 1 | 0
P. cymbiforme 0 0 0 1 0 1 0 0 0 0 | 1 dd 1 1 1 | 0
P. eucomum 1 0 1 1 0 |! 0 |! 1 0 |! 1 | 0 0 0 0 O0
P. goyasense 1 0 | 1 0 1 ол 0 0 D 1 1 1 0 1 0 0
P. heterotrichon 1 0 0 1 0 1 0 0 0 0 1 1 1 01 1 | 0
P. humboldtianum 0 0 0 +t 0 | 0 0 0 0 1 1 1 1 1 1 | 0
P. imbricatum 0 0 0 2 0 1 | B 0 0 ! 1 l 1 0 1 | 0
P. lanciflorum 1 1 0 | 0 | 0 1 0 1 lol 1 1 1 O ]|! O0
P. longiaristatum | 0 | | 0 O0 0 ] 0 O 1 1 1 O I 1 | |
P. malmeanum l 0 l l 0 1 0 0 | 0 l l 0 0 0 0 0 0
P. niquelandiae 0 0 0 1 0 I 0 0 0 0 D 1 0 1 0 0 I O0
P. pectinatum 1 1 1 1 O |! 0 1 0 0 1 1 1 1 1 1 I ©0
P. petrense 1 0 | 1 01 0 0 0 0 1 1 1 1 l 1 | 0
P. phyllorhachis 0 0 0 +t 0 ! 0 0 0 0 1 1 0 1 0 | | 0
P. polyphyllum 1 0 | | 0 1 0 0 0 0 1 1 0 1 l l | 0
P. reticulinerve 0 0 0 l 0 1 0 l 0 0 1 1 1 1 0 l l 0
P. stellatum 1 0 1 0 0 0 0 0 l 1 1 1 1 B 0 0 0 0
P. trachycoleon 0 0 0 l 0 l 0 0 0 0 l 1 0 1 0 l l 0
almum, P. cromyorrhizon, and P. notatum of group
Notata, and P. candidum, P. prostatum, and P. ra-
cemosum of group Racemosa were included in the
ма
present analysis (Table 4
Panicum obtusum Kunth and Anthaenantiopsis
fiebrigii Parodi were also included as outgroups due
to their relationship with Paspalum, stressed by a
recent molecular analysis of the tribe (Giussani et
al., 2001). Finally, for the purpose of rooting, Pan-
icum laxum Sw., a C, representative of the x = 10
clade of the Paniceae (Giussani et al., 2001) was
also considered in this analysis.
MORPHOLOGICAL CHARACTERS
Habit.
perennial, rarely annual (in P. longiaristatum and
Species of Paspalum subg. Ceresia are
Р. cachimboense). Culms are herbaceous, occasion-
ally lignified (in P. buchtienii and P. humboldtian-
um), unbranched or branched at the lower and mid-
dle nodes, hollow, and erect or leaning on the
vegetation. Blades range from filiform to linear or
lanceolate; a short pseudopetiole is present in P
buchtienti, P. humboldtianum, and P. cymbiforme,
while P. imbricatum is characterized by its long
pseudopetiolate blades.
Volume 89, Number 3 Denham et al. 345
2002 Paspalum subg. Ceresia
F
Figure 2. Scanning elec icu micrographs of the upper anthecium 9 —— —A. Р. pectinatum, apex of the upper
palea (Dusén 16057, SI). —B. ralmeanum. upper palea (Killeen 2024, SI). —C. P. heterotrichon, upper palea with
silica bodies and macrohairs — 4176, US). —D. Р. cordate, apex ol f the upper palea (Weddell 1699, P). —E. Р.
pee spikelet base, ventral view (Ailleen 247 1, SI).
US
| P heterotrichon, spikelet base, ventral view (Tovar 4176,
Annals of the
Missouri Botanical Garden
Inflorescences. Inflorescences are usually ter-
minal; axillary inflorescences are common in Pas-
palum buchtienii and P. polyphyllum. The number
of racemes varies from a single one, e.g., P. stel-
latum and P. carinatum, to two conjugate racemes,
e.g., P. malmeanum, P. eucomum, P. reticulinerve,
and P. pectinatum, while two alternate racemes are
characteristic of P. goyasense. When several ra-
cemes are present, these can be alternate and dis-
tant, e.g., P. trachycoleon, P. heterotrichon, P. cere-
sia, P. imbricatum, P. humboldtianum, | P.
polyphyllum, P. cymbiforme, P. aspidiotes, and P.
niquelandiae, or they can be subdigitate (with a
short internode, up to 1 em long, between the ra-
cemes) as in P. cordatum, P. lanciflorum, and P.
cachimboense. Pulvini are pilose, and one or two
bracts are usually present in P. stellatum, P. eucom-
um, P. goyasense, P. malmeanum, P. carinatum, and
P. pectinatum.
Rachis of the racemes. Species of Paspalum
subg. Ceresia are typically distinguished by a
winged, membranous to foliaceous rachis, 0.8—10
mm wide; the rachis is usually glabrous, occasion-
ally pilose (in Paspalum polyphyllum) or ciliate at
the margins (in P. reticulinerve); hyaline to mem-
branous, and nerveless margins are present in sev-
eral species, e.g., P. lanciflorum, P. cachimboense,
P. imbricatum, P. pectinatum, P. aspidiotes, P. het-
erotrichon, P. cymbiforme, P. ceresia, P. stellatum,
P. eucomum, P. malmeanum, P. carinatum, P. goy-
asense, P. biaristatum, and P. longiaristatum.
Although the rachis usually ends
point, in several
in a
species, such as P. humboldtian-
um, P. buchtienii, P. polyphyllum, P. niquelandiae,
and P burmanii, the rachis finishes in a naked
point or in a spikelet; in P. goyasense the rachis
always ends in a developed spikelet. Spikelets are
imbricate and arranged in two or four series.
Spikelets.
siventrally compressed, ranging from ovoid to ellip-
Spikelets are solitary or paired, dor-
soid to lanceolate. The size of spikelets varies from
1.4 to 8 mm in length; they are usually covered
with indument, otherwise ciliate; glabrous spikelets
are found in Paspalum phyllorhachis and P. imbri-
P. lon-
giaristatum and P. biaristatum. The lower glume is
always absent, while the upper glume is usually as
long as the spikelet and 3- to 5-nerved, 7(—9)-
nerved in P. aspidiotes and P. reticulinerve, occa-
sionally 2-nerved in P. stellatum.
catum. Awned bracts are present only in
Also, the upper
glume is cordate and winged in P. imbricatum, P.
reticulinerve, P. aspidiotes, P. cordatum, and P. pec-
tinatum. The lower lemma is winged in P imbri-
catum, P. reticulinerve, and P. aspidiotes. A lower
palea and lower flower are always absent.
Upper anthecium (Figs. 1, 2). The upper anthe-
cium is planoconvex, ellipsoid, ovoid or obovoid,
usually hyaline to membranous, but membranous
to chartaceous in Paspalum niquelandiae and Р.
goyasense. Margins of the upper lemma are flat, not
enclosing the apex of the upper palea. A short stipe
is present in P. stellatum, P. lanciflorum, P. cach-
imboense, and P. aspidiotes.
Regarding the upper anthecium ornamentation,
simple papillae regularly distributed, bicellular mi-
crohairs and macrohairs, more densely placed to-
ward the apex of the upper lemma and palea, are
common features of most species in subgenus Cer-
esia. A smooth and glabrous upper anthecium is
present in Paspalum malmeanum, P. eucomum, and
P. stellatum. Bicellular microhairs, 50-70 рт long,
have a short basal cell and an elongated apical cell.
Unicellular macrohairs, present or absent in sub-
genus Ceresia, can be short and hooked, ca. 40 jum
long, to long and thin, 60-300 рт long. Rounded
or elongated silica bodies are frequent in several
species; these silica bodies have usually 2 to 4 con-
strictions; 6 or more invaginations are present in P.
lanciflorum.
DisTRIBUTION AND. HABITAT
Species of Paspalum subg. Ceresia grow in
America, ranging from Mexico, Mesoamerica, and
the Caribbean (Haiti, the Dominican Republic, and
Trinidad and Tobago), to South America (Colombia,
Venezuela, Guyana, Surinam,
Peru,
Brazil, Ecuador,
Bolivia, Paraguay, central Argentina, and
Uruguay). Brazil hosts the largest number of spe-
cies, 22, and endemics, 10, in the subgenus.
Taxa are usually found in limestone, sandy, or
rocky soils, in open fields or mountain slopes, from
sea level to 3000 m elevation. Several species, such
as P. carinatum, P. lanciflorum, and P. pectinatum,
are found in open areas subjected to periodic fires;
these species have culms and sheaths adapted to
fire. Paspalum malmeanum and P. cordatum grow
in flooded savannas.
Paspalum stellatum, P. pectinatum, and P. cari-
natum are the most widespread species in subgenus
Ceresia. Paspalum stellatum grows from Mexico to
southern Brazil, Paraguay, northeastern Argentina,
and Uruguay: Paspalum pectinatum is found from
Mexico to southern Brazil, and P. carinatum from
Nicaragua to Brazil and Trinidad and Tobago. Re-
garding the biogeographical distribution of subge-
nus Ceresia, species are common in cerrados, At-
lantic forests, and savannas; they are also present
in caatingas of northeastern Brazil and in the Guy-
ana highlands of Guyana and Venezuela; they are
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
Table
shown in ion 4. with character
List of synapomorphies of the —
state changes in the
nodes and within taxa. Characters in bold are common
synapomorphies for all 40 most parsimonious trees.
Nodes 1-37
Node 1: character 55: 0 2 1, character 57: 0 > |
Node 2: character 54: 0 е l
Node 3: character 7: durs . character 14: 0 > I.
character 17: 0 > 1, chr ter 34: 0 > 1
Node 4: character 1 : , character 27: 2 > |.
character 28: | > 0, chara ter 43: 2 > 1
Node э: charac ter 40: 0 > l, character 51: 2 > 1,
character 52: 0 > 1, — 56:0 > |
Node 6: aie 9: 3 2
Node 7: charac id 24: 0 => l, character 26: 0 > 2
character 30: 1 > 2
Node 8 — 26: 0 > : character 50: 0 => |
Node 9: character 19: 0 = I. character 20: 0 > 1.
character 27: | > 2, athens 30: | => 0. od iter
31:0
Node 10: character 42: 0 > 1
Node 11: character 36: 0 = 1, character 40: | — 0
Node 12: character 2 > 0 2 1, character 27: 2 > 3.
e — ter 39:0 => 1
„ character 55: 0 > |
character 32: 0 =
Node 13: character A
| 4: character 4: =>
Node 15: character 46: 1 > 0, character 54:
Node 16: character 49: 0 = 1, character 56:
character 57: | => 0
Node 17: character 26: 2 > |,
character 30: 2 > 3
Node 18: character 45:
] e
] 260,
character 29: 1 > 0,
1 = 0, character 46: y >3
— 0, charac Е 21:05 I,
character 38: 0 = 1. бт ви ter 48: bri
Node 20: character 18: | > 0
Node 21: character 13: : > 1. character 16: 0 > |
Node 22: character 10:
Node 23: character T ! = 0, character 53: | > 0.
character 55: | >
character 40: | => 0
Node 24: character * 02 1.
| : | => 2, character 25: 0 > |
: character 15: 2 > 1,
| character 18: 0 > 1
Node 29: character 1: 0s I. >
character 12: 0 > |
Node 31: character 28: 0 2 1, character 44: 0 > |
Node 32: character 9: 3 > 2, character 30: 1 > 2.
— tet A
Node 33
|
| . character 3: 0 > |
Node 36: character 24: 0 = 1. character 26: 0 > 3,
:h l, character 42: 0 => 1
Node 37: character 43: 2 = 1, character 44: | 0
Outgroup taxa
Panicum obtusum: character Ө: 1 — О, character 27:
233
ые кл, fiebrigii: — ter 24: Ө — 1. charac-
r 26: 0 > 3. chi — => 0. — 10: 0
> l, character 42: ( : не ter 51: 2 > 1,
character 52: 0 — |
Paspalum cromyorrhizon: character 4: 0 — |
P. notatum: no autapomorphies
Table 3. Continued.
P. almum: character 10: 0 > 1, character 57: | = 0
P acuminatum: character 30: 1 > 0. character 35: |
> 0, character 46 =>» 0, * ter 56: Ө => |
P. dissectum: oneracios 10: 1 = 2, character 49: 0 > 1.
character € = 0
P. sanguinolentum: character 15: e
P. erianthum: character 15: 0 > 1. character 18:
0. character 35: | — 0, — дег 55: 1 =>
Г. — character 4: 0 = 1, character 46:
char 54:12
R ee character 55: 1 > 0
P candidum: character 35: | > Ө, character 46:
P. racemosum: character 54: | =>
12690.
] ze
Ingroup taxa: Subg. Ceresia
P. lanciflorum: character 56: | = 0
P. cachimboense: character 1: 0 — 1,
1 = 0
P. pectinatum: no ene PAPE: 5
P. cordatum: character 4: А 2a ter 10: 3 > 1
P. imbricatum: character 5: 0 = 2
P. reticulinerve: character 10: 3 > 1,
01
P. aspidiotes:
P. goyasense: character 11:
| e
P. carinatum: character 48: 0 > |
P. eucomum: character 48: 0 > 1
P. malmeanum: character 53: | > 0
P. ceresia: ае 35:1 > 0, character 51:
character 53: 0. PM E r55:0 =
stellatum: characte ‘ter 49; ( . charac ter 50: 0 > 1.
О. AN => 0
0 = 1. character 55:
character 55:
character 15:
character 9: 2 => 3. character 50: 0 => 1
1 > 0, character 5
] шә 0,
Р
character 56:
P. longiaristatum: habitos l:
0s]
P. biaristatum: no auté pop S
Р S terotrichon: character 26:
=> 0
P — character 42: 0 >
P. trachycoleon: no mgro an 5
Р. жү жө E chang : 1 0. character 26: 2
charact
Р, Bd pa no a йан ы ТРИ
P. buchtienii: no autapomorphies
P. humboldtianum: character 26: 2 —
p د character ч @ а.
= 1, character 42: 0 =
р ОЛ? character 26 2 => 0, character 55:
). character 56: | => (
. character 54:
3
3, character 40: 0
less frequent in the Amazonian forests. Several spe-
cies have an Andean distribution: P. humboldtian-
um. from Mexico and Mesoamerica to northwestern
and central Argentina; P. buchtienii is restricted to
mountain slopes of Bolivia and Peru, while P. cym-
biforme is found in Mexico and Mesoamerica. A
disjunct ынана was observed in the following
species: Р. polyphyllum, present in southern and
central Brazil, in the states of Rio Grande do Sul
and Santa Catarina, southern Paraguay, northeast-
ern Argentina, and Uruguay, and also in Colombia
348 Annals of t
Missouri — Garden
Table 4. List of species of informal groups of Paspalum Eriantha, Dissecta, Notata, and Racemosa used as outgroups
in the cladistic analysis and species of Paspalum subg. Ceresia and its sections Ceresia and Pectinata.
Outgroups
Dissecta Group (sensu Morrone et al., 1996) Paspalum а
? dissectun
Eriantha Group (sensu Chase. Ined.) P. erianthum
P. guttatum
P. sanguinolentum
Notata Group (sensu Canto-Dorow et al., 1996) P. almum
P. cromyorrhizon
P. notatum
Racemosa Group (sensu Morrone et al., 1995) P. candidum
P. prostatum
P. racemosum
Paspalum subg. Ceresia section Pectinata P. aspidiotes
P. cachimboense
F. cordatum
P. imbricatum
section Ceresia
P. lanciflorum
P. pec tinatum
P. reticulinerve
Р. —
P. buc
P. —
P. carinatum
P. ceresia
P. cymbiforme
P. eucomum
1sense
P i hon
P. humboldtianum
P. longiaristatum
P. malmeanum
P. niquelandiae
P. trachycoleon
and Venezuela. Paspalum cordatum has been col-
lected in southern Brazil, Amazonia, and Colombia.
Paspalum ceresia is common in the Andean region
from Ecuador to northwestern Argentina, and iso-
lated populations grow in central Brazil, in the
states of Goiás, Maranhao, Minas Gerais, and Para.
Finally, P. trachycoleon grows from Mexico and Me-
soamerica to Colombia and Venezuela and reap-
pears in southern Brazil.
As previously mentioned, a significant concen-
tration of species is found in the cerrados of central
Brazil. Paspalum biaristatum, P. longiaristatum, P.
burmanii, and P. niquelandiae are restricted to ser-
pentine soils of the state of Goiás; P. petrense, P.
imbricatum, and P. goyasense are endemic species
of this state; P. cachimboense was found only in the
on the
Serra do Cachimbo, state of Mato Grosso.
border with the state of Рага, and P. phyllorhachis
in Minas Gerais. Another species of subgenus Cer-
esta endemic to central Brazil is P. eucomum, pres-
ent in Minas Gerais, Sáo Paulo, Paraná, Goiás, and
in the Distrito Federal. It is noteworthy to mention
this high number of endemic species in serpentine
soils, characterized by high concentrations of nick-
el. chromium, cobalt, and other minerals (Brooks,
1987), combined with a low percentage of nutrients
in these soils. Other endemic grasses were also pre-
viously described for this region (see Filgueiras et
al., 1993, for a detailed analysis of the area).
PHYLOGENETIC ANALYSIS
The cladistic analysis using NONA resulted in
40 most parsimonious trees, each of 196 steps,
found in 84 of 100 replications, with CI = 38 and
RI = 68. Searches to find possible islands yielded
no additional trees.
Volume 89, Number 3 Denham et al. 349
2002 Paspalum subg. Ceresia
Panicum laxum
Panicum obtusum
E Anthaenantiopsis fiebrigii
Paspalum erianthum
P. guttatum | Eriantha Group
P. sanguinolentum
1 P. acuminatum .
| Dissecta Group
1 P. dissectum
P. cromyorrhizon
n P. notatum | Notata Group
— — —1
> P. almum
75 Р. candidum ;
| Р. racemosum ; Racemosa Group
A P. prostatum
P. lanciflorum 7]
P. cachimboense
P. pectinatum
-
2 Р. cordatum a Y
22
P. imbricatum Fd ;
Р. reticulinerve
Р. aspidiotes =
P. goyasense =
2 Р. сагіпаќит
Р. сегеѕіа
Р. stellatum
-
P. eucomum » B
ов
Р. таітеапит a =
Ф
Р. longiaristatum 5
c
l 3 P. biaristatum &
P. heterotrichon © Ф |
— —— Р. petrense 2 & |
P. trachycoleon |
1 Р. phyllorhachis
Р. cymbiforme
[99]
63 Р. buchtienii
Р. humboldtianum
Р. polyphyllum
Р. niquelandiae
1 ii | |
Р. burmanii ال d o
Figure 3. ct consensus tree from 40 most parsimonious c — ams found by NONA. Numbers below branches
indicate Bremer E numbers above branches indica
Figure 4 shows one of the 40 most parsimonious
trees. Unambiguous synapomorphies of each node
are listed in '
state change occurs just in that cladogram or in all
equally parsimonious trees.
З. with
Bremer support values and jackknife frequencies.
The strict consensus is shown in Figure
In this consensus tree, Paspalum is monophyletic
with Anthaenantiopsis fiebrigii or Panicum obtusum
as sister taxa. Next within Paspalum, there is a
Table 3. which indicates if a character
alues.
te jackknife
major clade including the Racemosa group and
subgenus Ceresia. Relationships of the remaining
studied species of the genus, of Eriantha, Notata,
and Dissecta groups, are not solved; they are placed
in a basal polytomy in the tree. Two different to-
pologies are equally parsimonious when solving this
polytomy. according to the placement of species of
the group Eriantha.
The clade comprising groups Racemosa and Cer-
esta (Fig. 4, see node 4) is supported in all most
350
Annals of the
Missouri Botanical Garden
Panicum laxum
Panicum obtusum
Anthaenantiopsis fiebrigii
P. prostatum
igure 4.
parsimonious trees by character 11 (apex of the ra-
chis ending in a naked point) and character 27 (up-
per glume 3-nerved). In some of the most parsi-
monious trees, character 28 (nerves of the upper
glume equidistant) and character 43 (lower lemma
3-nerved) also support this clade.
Paspalum subg. Ceresia forms a clade with Ra-
cemosa as ils sister group. At node 5 (Fig. 4), char-
acters supporting the monophyly of subgenus Cer-
esta are: 51 (membranous upper anthecium) and 52
P. acuminatum
P. dissectum
P. sanguinolentum
P. candidum
Р. racemosum
P. cachimboense
P. goyasense
One of the 40 most parsimonious trees found by NONA (196 steps le ын Cl =
number of nodes in boxes. Synapomorphies of each node and taxa are listed in Table
|
| Notata Group
|
P. almum |
Dissecta Group
|
|
|
P. erianthum | Eriantha Group
P. guttatum |
| Racemosa Group
P. lanciflorum
P. imbricatum
o
9
a
E
Se
3
ejeunoeg әс̧
P. reticulinerve
P. aspidiotes
P. carinatum
P. eucomum
P. malmeanum
P. ceresia
P. stellatum
еѕәзәо ‘Bans unjedseg
P. longiaristatum
P. biaristatum
P. heterotrichon ‚
Р. petrense |
eIsaja) ‘Pes
Р. trachycoleon
Р. phyllorhachis
Р. cymbiforme
Р. buchtienii
Р. niquelandiae
P. burmanii
38, RI = 68) with
(upper palea gaping at the apex). Characters 40
(lower lemma with pilose margins) and 56 (upper
lemma with unicellular macrohairs) also support
subgenus Ceresia in some of the most parsimonious
trees.
Within Paspalum subg. Ceresia, two major clades
appear in the analysis: one containing species of
section Pectinata, and the other containing the re-
maining species of the subgenus. Section Pectinata
(see node 6, Fig. 4) is supported by character 9
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
P. stellatum
P. eucomum
P. malmeanum
P. ceresia
P. longiaristatum
P. biaristatum
#9, 49, 56, 57
Figure 5.
Figure 4. —B. Node 20 of cladogram of Figure
(presence of approximate racemes); within this sec-
tion, P. pectinatum, P. cordatum, P. imbricatum, P.
reticulinerve, and P. aspidiotes are grouped together
and are very well supported by the presence of
characters 19, 20, 27, 30, and 31 (upper glume
winged, cordate, 5-nerved, with flat margins, and
marginal nerves not reaching the apex) in all of the
most parsimonious trees
Fig.
cluding species here considered to be in section Cer-
The other main — (see node 7, 1). in-
esia, is supported, in all shortest trees, by characters
24 (upper glume with pilose margins). 26 (upper
glume hairy on the lower half), and ЗО (margins of
the upper glume folded at the nerve). Within this
clade, two subgroups (Figs. 3, 4) are recognized:
clade A, including Р. goyasense, P. carinatum, P.
eucomum, P. malmeanum, P. ceresia, P. stellatum, P.
longiaristatum, and P. biaristatum, with the following
characters as synapomorphies (node 13, Fig. 4): 42
(lower lemma pilose) and 55 (upper anthecium with-
out silica bodies). In all of the most parsimonious
trees, P. longiaristatum and P. biaristatum constitute
a well-supported terminal group (node 19. Fig. 4).
both species with an awned upper glume and lower
lemma (chars. 21 and 38). Clade B includes P. het-
erotrichon, P. petrense, P. trachycoleon, P. phyllor-
hachis, P. cymbiforme, P. buchtienii, P. humboldtian-
um, P. polyphyllum, P. niquelandiae, am
burmanii, and is defined by character 18 (hyaline
upper glume) in all trees (node 20, Fig. 4).
Two equally parsimonious topologies were found
in clade А: one shown in Figure 4 (node 18), where
P. stellatum is the sister species of P. longiaristatum
P. heterotrichon
P. petrense
P. trachycoleon
P. phyllorhachis
P. cymbiforme
#40 P. buchtienii
P. humboldtianum
P. polyphyllum
P. niquelandiae
P. burmanii
Alternative most parsimonious topologies within Paspalum sect. Ceresta.—A. Node 15 of cladogram of
and P biaristatum, due to support from characters
15 and 46 (lower lemma with marginal nerves and
extension of the margins absent). In the other reso-
5A, P. stellatum,
um, and P malmeanum constitute a group supported
lution, shown in Figure P. eucom-
by characters 9 (two racemes per inflorescence). 49
(obovoid upper anthecium), and 56 and 57 (upper
anthecium without macro- and microhairs), with P
ceresia as the sister species of P. longiaristatum and
P. biaristatum.
On the other hand, two equally parsimonious to-
pologies were obtained for clade B: in one of them,
shown in Figure 4, P heterotrichon, P. petrense, P.
trachycoleon, and P. phyllorhachis are monophyletic
(node 21), being sustained by characters 13 (spike-
lets laterally inserted in the pedicels) and 16 (spike-
lets with an annular The
alternative topology (Fig. 5B) places P trachycoleon
and P. phyllorhachis as sister group to a clade con-
sisting of P. cymbiforme, P. buchtienii, P. humbold-
thickening at the base).
tianum, P. polyphyllum, P. niquelandiae, and P. bur-
manii. This clade. with eight species, shares a
elabrous margin of the lower lemma (char. 40).
Two most parsimonious trees (fit = 413.7 and 199
steps) were obtained using implied weights with
PeeWee (trees not included). Regarding the Paspal-
um ingroup, the strict consensus of both trees has
the same topology as the tree, obtained with NONA
(Fig. 3). where group Racemosa is the sister to sub-
venus Ceresia. Characters supporting clades in the
two shortest trees are the same as in the equal weight
analysis, except character 24 for section Ceresia
4). In both most parsimonious trees
(node Fig.
352
Annals of the
Missouri Botanical Garden
from Pee Wee, groups Eriantha, Dissecta, and Notata
are more closely related to each other than to the
clade group Racemosa + subgenus Ceresia.
DISCUSSION
Monophyly of Paspalum subg. Ceresia is sustained
by the present analysis. Two additional steps are re-
quired for breaking down the subgenus Ceresia
clade, which has a Bremer support of 2 and a jack-
knife value less than 50; similar low values are com-
mon in different cladistic treatments using morpho-
logical data (Karis, 1995). Only homoplasic
characters support this clade including subgenus
Ceresia: a membranous upper anthecium (char. 51)
is present in nearly all species of the ingroup, but
P. niquelandiae and P. goyasense have a membra-
nous or chartaceous upper anthecium, whereas in P
ceresia it is hyaline. Regarding the outgroups, the
upper anthecium is membranous in Anthaenantiopsis
fiebrigii, and membranous to chartaceous in P can-
didum and P. cromyorrhizon. Character 52 (upper
palea gaping at the apex, not covered by the upper
lemma) is a synapomorphy of the subgenus but also
present in A. fiebrigii. The winged rachis, widely
used to recognize the subgenus, is present in many
of the outgroup taxa, evidently arising many times,
or being lost, in the evolution of the genus.
Within subgenus Ceresia two monophyletic sec-
tions are recognized: section Peetinata, including
Paspalum lanciflorum, P. cachimboense, P. pectin-
atum, P. cordatum, P. imbricatum, P. reticulinerve,
and Р. aspidiotes, and section Ceresia, including
P. goyasense, P. carinatum, P. eucomum, P. mal-
meanum, P. ceresia, P. stellatum, P. longiaristatum,
P. biaristatum, P. heterotrichon, P. petrense, P. tra-
chycoleon, P. phyllorhachis, P. cymbiforme, P. buch-
tienii, P. humboldtianum, P. polyphyllum, P. nique-
landiae, and P. burmanii. Section Pectinata is
weakly supported herein (Bremer support 1;
jackknife value less than 50); a similar delimitation
of section Pectinata was recently shown by Rua and
Aliscioni (in press). Within this section, the group
including P. pectinatum, P. cordatum, P. imbrica-
tum, P. reticulinerve, and P. aspidiotes is well sup-
ported (Bremer support = 4; jackknife value = 81),
all species sharing a winged and cordate, 5- or
more nerved upper glume, with the marginal nerves
not reaching the upper third of the glume. Paspal-
um lanciflorum and P. cachimboense are basal spe-
cies within section Pectinata, both taxa without a
winged and cordate upper glume, but with a lower
lemma pilosity similar to other species of section
Pectinata.
Section Ceresia is also weakly supported herein
(Bremer support = 1; jackknife value less than 50).
Within the section, Paspalum biaristatum and P
longiaristatum are well-supported sister species
(Bremer support = 3; jackknife value = 91) but
are not distinct from section Ceresia. Therefore, we
consider section Biaristata (Filgueiras & Davidse,
1994) as included in section Ceresia, because re-
taining section Biaristata within our section Ceresia
will render section Ceresia monophyletic.
TAXONOMIC TREATMENT
Paspalum subg. Ceresia (Persoon) Reichenbach,
Consp. Regn. Veg.: 49. 1828. Ceresia Pers.,
Syn. РІ. 1: 85. 1805. Paspalum sect. Ceresia
(Pers.) Nees, Fl. Bras. Enum. Pl. 2: 76. 1829,
[as "Paspalus sect. Ceresiae"]. Paspalum sect.
Ceresia (Pers.) Schltdl. ex Müll. Stuttg., Bot.
Zeitung (Berlin) 19: 326. 1861. nom. —
Paspalum subsect. Ceresia (Pers.) Benth., Gen.
3: 1098. 1883. Paspalum sect. Ceresia
(Pers.) Pilg., Repert. Spec. Nov. Regni Veg.
26(15): 230. 1929, nom. illeg. TYPE: Paspal-
um ceresia (Kuntze) Chase [= Ceresia elegans
Pers. ].
PH sect. d F Es iras & Davidse, Novor
1994. Syn. YPE: Paspalum cin
le & Davidse
Paspalum sect. Pectinata Chase ex Rodr.-Rodr.,
Ernstia 2(1-2): 22. 1992. TYPE: Paspalum
pectinatum Nees ex Trin.
Plants perennial, erect or leaning on the vege-
ation. Blades rigid, filiform to lanceolate. Inflores-
cences with 1 to several racemes, rachis of the ra-
—
cemes winged, hyaline to membranous, 0.8-10 mm
wide. Spikelets 1,4—8 mm long, pilose, occasionally
glabrous; lower glume absent; lower palea and low-
er flower absent: upper anthecium pale, hyaline to
membranous, chartaceous in some species, the pa-
lea free at the apex.
‘Twenty-five species growing from Mexico to Ar-
gentina, Uruguay, and southern Brazil.
Cerea Schltdl., Bot. Zeitung (Berlin) 12: 820.
1854, is a superfluous name for Ceresia.
KEY TO DISTINGUISH THE SPECIES OF
PASPALUM SUBG. CER
l. Upper a НР | 2
1". Ipper glume not winged 6
2(1). Spikelets glabrous 14. P imbricatum
2: Spikelets pilose 3
3(2). Spikelets paired, arranged in four series
3. P. retic iculinen е
3. Spikelets solitary, arranged in two series 1
4(3. Lower lemma she nd ciliate on the maris,
not papillose-pilos 1. P. aspidiotes
4”. ie wer lemma денш uously papillose- тов se
the margins ме, 5
5(4). Cite short rhizomatous, 35-100 cm tall;
sheaths red or brown, shiny at the outer sur-
face; racemes 2(3 to 5), 2-8 em long; spike-
Volume 89, Number 3 Denham et al 353
2002 Paspalum subg. Ceresia
er lanceolate, 5-8 mm long, 2-3.3 mm Б Perennial; upper anthecium not — 16
Wide aaa 19. P. — 16(15). Spikelets solitary; the margins not corky,
5, Culms with long, curved rhizomes, 100— with hairs up to 3 mm — appressed or
cm tall: sheaths stramineous or green, not ascendent, irregularly distributed on the
shiny at the outer surface; racemes 5 to es margins í
12-16 cm np spikelets ovoid, 4—5(—6) n lo’. еј pi aired, occ asionally the lower one
long, 2.5-3.5 mm wide „Р can — ed or aborted; the margins corky, with
6(1) Spikelets with the upper glume and the lowe radiate marginal hairs (ascendent in P cym-
lemma awned — А d biforme) 20
Y. Spikelets not awned 0 8 170106). Racemes 2, conjugate; spikelets .1.6-3 mm
1(6) Annual, culms 15-36 « m tall: spikelets 1. E long. the margins densely pilose on the up-
2.2 mm long; awn of the upper glume 6-12 per portion, with hairs up to 2 mm long.
mm long. that of the lower lemma 0.3-2 1 shortly pilose toward the base: upper anthe-
long : 16. P. P longiaristatum cium as long as the upper glume and lower
T! Perennial, culms 70-125 cm tall: spikelets lemma We a H 18
3.8-4.5 mm long; awn of the upper У 1 17’. Racemes 1 to 3, not conjugate: spikelets 3—
7.1 mm long, that of the lower 6 mma 3.8 5 mm * with long hairs on the basal mar-
1.5 mm long |... . P. biaristatum gins, sparsely and shortly pilose on the upper
8(0) Rachis (3.5-)4—10 mm wide __ © 9 portion, occasionally glabrous; upper anthe-
8”. Rachis 1-2.5-3) mm wide 15 cium shorter than the upper glume and lower
9(8) Spikelets solitary — — 10 lemma 19
o". Spikelets — 13 18017). Spikelets | 6-2 mm long 17. P. malmeanum
10(9). Spikelet (4.8—)5 6-7. 2 mm long; anthiecium 18. Spikelets 2.5-3 mm long 10. P. eucomum
№ to Y length of the spikele 19(17). Racemes l, occasionally 2: rachis of the
= 15 P. ыа branches 1.8-2.5(-3.5) mm wide. straight to
10’. Spike let 2.3-3.9 mm lane: аа ‘ium as long subfalcate; upper glume densely pilose on
as or slightly shorter than the spikelet 11 the lower third, ra scaberulous: —
11(10). Inflorescences with l (rarely 2) racemes, if 2 glume and lower lemma flat in the upper por
conjugate; pulvini with | or 2 papyraceous tion: blades filiform 6. P. carinatum
bracts: pedicels with a crown of hairs toward 19”. Racemes (1)2(3), rae dis e the branches 1—
the apex: upper anthecium shortly stipitatez 1.6 mm wide, straight; lower glume densely
blades filiform 24. P. stellatum pilose in the lower half, otherwise glabrous;
П”. Inflorescences with 1 to 7 alternate racemes: upper glume and lower lemma slightly con-
pulvini ebracteate; pedicels hispid, without a dades linear Е 11. P. govasense
crown of hairs toward the apex: upper anthe- 20. Rac iua of the branches 2-3 mm wide 2]
cium not stipitate; blades linear to linear-lan- 20". Rachis of the branches 0.8-1.541.8) mm
eolate (occasionally filiform in P ceresia) 12 wide 277 22
12(11). ie his 3.54 mm wide; blades 1.5—1 mm 21(20). — ‘ences 8-12 em Jong, with і ra-
wide; spikelets 2,3—3 mm long, not early de- emes; floriferous culms D. em tall:
ciduous, with an annular thickening at the blades linear, 5-19(-24) x 0.2—0.5(—0.8)
base: upper glume with papillose-pilose em: pseudoligule a ring of hairs
hairs toward the — a few of them lon- 0 —— — —— 9. P. cymbiforme
ger than the rest; lower lemma with long 21^. Inflorescences 20-25 cm long, with 9 to 11
hairs at the upper margins, otherwise & racemes; floriferous culms 150 cm tall;
brous: lodicules : 12. P. НАУ сее hon blades * ‘eolate, 30-35 X 1-1.2 em; pseu-
12’. Rachis 6-10 mm wide blades 3-15(220) doligule absent - v 4. P. burmanti
mm wide; spikelets (333.4-3.6 mm long. 22(20). Plants — 120-185 em tall, blades 20—
early a without an annular thick- 15 0.8-2.2 ст, inflorescences with 0
ening al the base; upper glume with + equal. ae racemes .....................-- 18. Р —
silky, not papillose-pilose, marginal hairs: 22'. Plants (ode 'eous, 30-100 cm tall; blades
lower lemma ee all over the surface ez lod- 1-21 em long, 0.1-1.8 cm wide, inflores-
icules ahsent P. ceresia cences with | to 11 racemes 2:
13(9). Plants нта spikelets зн 23(22). Rachis pilose on one or both surfaces; blades
2]. кт n * аг to linear-lanceolate, 4-13 cm long,
13’. Plants herbaceous: apie lets pilos .1—0.6 em wide; upper glume with une ja
14(13). Culms unbranched, 70-90 cm tall: — "s 3 marginal hairs; lower lemma with —
6 mm wide; rachis (6.5—)7—7.5 mm wide: up- cal margins |... 22 P plum
per glume long-acuminate; lower li mma pi- 23"; Rachis дарш blade 's lanceolate, 5-21 cr
lose on the — 'r half, not sulcate long. 0.5-1.8 cm wide; upper Jin =
— 20. P. mum marginal hairs evenly distributed’ all of the
14”. Culms ене hing, — m tall: blades 0.5-1.5 same length; lower lemma with glabrous api-
cm wide; rachis 4-6 mm wide; upper pus cal ma <A, ин 2
acute to acuminate; lower lemma pilose at 24(23). Racemes (4)7 to 11. arcuate: rac his of the
the upper pd or — sulcate еи 0.8-1.2(-1.4) mm wide; spikelets
— 25. P. trachycoleon 0.8-1(-1.3) mm wide, sparsely distributed;
15(8). Kal upper anthecium long-stipitate blades papyraceous |... 3. P. buchtienii
— 5. P. cachimboense — 24". Racemes 2 to 6, not arcuate; rachis of the
Annals of the
м Botanical Garden
354
branches 1.2-1.5 mm wide; spikelets
1.4 mm wide, imbricate; ou herbaceous
. P. humboldtianum
1. Paspalum aspidiotes Trinius, Sp. Gram. 3: pl.
269. 1829-1830. TYPE: Brazil. Without lo-
cality: С. Н. von Langsdorff s.n. (holotype, LE!
photo, SI!; isotypes, P!, US-2942148!). Figures
6, 7.
Ж арш C ic Brittonia 3: 150, fig. 1. 1939.
TYPE: Ven 2 la. Bolívar: a tepuf,
1100 m, “Die. 1937-Jan. 1 G. H. I ^ 1265
— NY not seen, КІ; — MO-
2704419!, US-172: 362: 3! Fa SI!).
е ‘retin Swallen, Fieldiana, Bot. 28: 22.
TYPE: Venezuela. Bolivar: Gran Sabana, bor-
-
de Kavanayén and base of Ptari-tepuf,
Nov. 1944, J erman ‹ 60302 (holotype, US-
1911668! photo, SI!; isotypes, MO-3001316!, NY
not seen, VEN not seen).
Rhizomatous perennial, with arcuate rhizomes;
floriferous culms 30-90 cm tall, 0.3—0.4 cm diam.,
erect, simple or branched; internodes cylindrical,
glabrous, finely striate; nodes compressed, shortly
pilose, brown. Sheaths usually longer than the in-
ternodes, overlapping, densely villous all over the
surface or only at the distal portion, the margins
pilose. Ligules membranous, 3—4 mm long, ferru-
ginous, glabrous; pseudoligule a tuft of long white
hairs. Blades lanceolate, 8-32 cm long, 0.7-1.5 em
wide, attenuate at the base, the apex acuminate,
rigid, ascendent, densely villous on both surfaces,
the margins pilose. Terminal inflorescence exsert-
ed, subdigitate. Peduncles up to 30 cm long, terete,
finely striate, glabrous; main axis up to 4 cm long,
flattened, glabrous or sparsely pilose; pulvini pi-
6-13.5
cm long, alternate; rachis of the racemes flattened,
lose; racemes 4 to 7, rarely 3, ascendent,
34.5 mm wide, ending in a naked point, shortly
pilose, the midnerve manifest, the margins mem-
branous, pale to brown; pedicels short, flat, pilose
toward the apex; spikelets solitary, densely imbri-
cate in 2 series. Spikelets broadly ovoid, 6.5-7.5
mm long, 3—4 mm wide, dorsiventrally compressed,
pilose, ferruginous, upper glume and lower lemma
winged, cordate, the apex rounded. Upper glume
papyraceous, 7(-9)-nerved, with 3 central nerves
reaching the apex and anastomosed veins between
the nerves, with 1 to several white hairs, up to 2.5
mm long, at the apex, the rest of the surface short
ciliate, the margins ciliate. Lower lemma glumi-
form, 5.5 mm long, 3.1 mm wide, papyraceous, 5—
T-nerved, corrugate toward the basal portion, cili-
ate, with one to several white hairs, at the upper
margins, otherwise scabrous. anthecium
Upper
long-ellipsoid, 5 mm long, 1.3 mm wide, shorter
than the spikelet, membranous, stramineous, plano-
convex, shortly stipitate, minutely papillose, with
simple papillae all over the surface, scabrous and
with a tuft of macrohairs, bicellular microhairs, and
silica bodies at the apex; lodicules 2, conduplicate,
hyaline, 0.4 mm long; stamens 3, anthers 1.6 mm
long; stigmas 2, plumose, lateral. Caryopsis not
seen.
Distribution and habitat. Paspalum aspidiotes
grows in savannas, on dry and rocky soils, between
100 and 1200 m elevation in Venezuela (Amazonas
and Bolfvar), Colombia (cited by Judziewicz, 1990),
Guyana, Brazil (Amazonas), and Bolivia, where it
was recently found in Santa Cruz.
Paspalum aspidiotes is related, within section
the latter. distin-
guished by paired spikelets. It is also related to P.
cordatum Hack., a species with spikelets 4—5(—6)
mm long, without white hairs on the apex of the
upper glume and lower lemma, the upper glume
and lower lemma not scabrous over the whole sur-
Pectinata, to P. reticulinerve,
face, and with conspicuous marginal, tuberculate
hairs at the lower lemma
The study of type material of P. aspidiotes, from
LE, and that of P. setiglume, from K, MO, and US,
allowed us to verify that there are no differences
between the taxa. There is, within the examined
specimens, variation in the number and length of
the hairs on the upper margins of the upper glume
and lower lemma. Therefore, Paspalum setiglume is
considered a synonym of P. aspidiotes.
The specimen Riedel s.n., collected in Brazil
without a definite locality and examined at P, prob-
ably represents an isotype of P. aspidiotes (see Ren-
voize, 1978, for comments on Langsdorff and Riedel
collections).
Repres sentative specimens. BOLIVIA. Santa Cruz: Ve-
. Amazonas: Estrada «
road to Igarapé Preto, Galdaran et al. 2760 (US);
Savanna, i Puciari tributary of Rio Ituxi, Prance et al.
13823 ү» IS); Mun. Hurai — Humaitá-Jacare-
canga, exeira et al. O). Without state:
without ia ality, Riedel 1048 e Bu s.n. (G, P). GUY-
ANA. Imbaimadai savanna, Upper Mazaruni River, Ay-
е Plateau, 370 (US). VENEZUELA. Ama-
onas: Territorio — Amazonas, 4
— 3618 (MO). Bolivar: M Gran Sabana, km 145
alo ong hwy., و (K, US); Distr. Piar, at top
› Aicha near E base of ee -tepui, Davidse &
44 (MO); Dist. Rosci чо, 5 de Cantarana, Huber
cabeceras no-
rorientales del Río Urimán, Huber et al. 9898 (MO, US);
by first ford W along summit Torom-Meru NW of Parupa,
Kral 72050 (MO).
Fr Taser .
^
©
Volume 89, Number 3 Denham et al. 355
2002 Paspalum subg. Ceresia
Figure б. Paspalum aspidiotes, —А. Habit. —B. Detail of ligule. —C. Portion of the rac pos —D, К, F. Spikelets,
uppe r glume view. —G. Spikelet, lower lemma view. —H. Upper anthecium, dorsal view. —I. U} anthecium, ventra
view. —J. Upper palea, lodicules, and gynoecium. (A-C, E-J based on Riedel s.n.. Р: D base di on — 1285, US
356
Annals of the
Missouri Botanical Garden
—
A Р оо
ө Р aspidio
xP —
о 200 400 600 800 1000 km
0 100 200 300 400 500 600 miles
Prepared by Hendrik R. Rypkema
100
Figure 7.
2. Paspalum biaristatum Filgueiras & Davidse,
Novon 4: 18, fig. 1. 1994, TYPE: — Goiás:
Niquelándia, Macedo, ca. 14°18'S, 48°23' W.
18 km N de Niquelândia, 10 June i T. S.
Filgueiras 2341 (holotype, IBGE!; isotypes, B
not seen, F not seen, FLAS not seen, ICN!,
ISC not seen, K! photo, SI!, MEXU not seen,
MO!, P not seen, R not seen, RB not seen, SI!,
SP not seen, UB not seen, US!). Figure 7.
Caespitose, shortly rhizomatous perennial; culms
simple or branched, 70-125 cm tall, 1—1.3 mm
internodes cylindrical, hollow, gla-
diam., rigid;
brous; nodes dark, the basal ones glabrous or pi-
lose, the upper ones glabrous. Sheaths 4—7 cm long,
shorter than the internodes, keeled, with papillose-
pilose hairs, the margins glabrous or ciliate. Ligules
membranous-ciliate, the membranous portion 0.2—
0.3 mm long, cilia 0.2-0.3 mm long; pseudoligule
a ring of hairs 1-1.2 mm long. Blades linear to
linear-lanceolate, 7-19 cm long. 0.1—0.9 em wide,
7. Distribution of Paspalum aspidiotes, P. biaristatum, and P. polyphyllum.
L potyptv
papillose-pilose on both surfaces, attenuated
rounded at the base, the apex long-acuminate, the
margins ciliate. Peduncles subexserted or exserted,
up to 17 cm long, terete. Terminal inflorescences
6-11 cm long, 3—4 cm wide; main axis absent or
up to 6 em long, glabrous, scabrous; racemes 1 to
4(7), 3—8(-13) cm long, solitary or alternate, dis-
tant, arcuate, ending in a naked, acuminate point;
pulvini pilose; rachis of the racemes 4.5-7 mm
wide, glabrous, with a midnerve prominent and
membranous, sparsely ciliate; pedicels
margins
mm long, solitary, hispidulous; spikelets
0.3-0.5
densely imbricate in two series along the rachis.
Spikelets long ellipsoid, awned, 3.8-4.5 mm long
without the awn, 0.4-0.5 mm wide, pilose, strami-
neous to purple. Upper glume hyaline, 3-nerved,
densely pilose toward the base, with hairs reaching
%—%5 the length of the spikelet, otherwise glabrous,
the margins densely pilose with hairs 1.7-2.2 mm
long: awn 4—7.1 mm long, scabrous. Lower lemma
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
slightly shorter than the upper glume, membranous.
3-nerved; terminal awn 3.8—4.5 mm long, scabrous.
Upper anthecium ellipsoid, 3-3.5 mm long. 0.5 mm
wide, shorter than the spikelet, papyraceous, shiny.
pilose on the upper portion: upper lemma 5-nerved,
the nerves inconspicuous, with bicellular micro-
hairs and macrohairs toward the apex, the margins
hairy on the upper third or glabrous, flattened: up-
per palea with microhairs, macrohairs, hooks, and
silica bodies on the distal portion; lodicules absent:
stamens 3, anthers 2—2.2 mm long; stigma plumose.
Caryopsis 1.2-1.5 mm long. 0.3-0.5 mm wide: hi-
lum punctiform, embryo less than half the length
of the carvopsis.
Distribution and habitat. Endemic to central
Brazil: it grows in Niquelándia, Goiás, in serpentine
soils at 1000 m.
Paspalum biaristatum is related to P. longiaris-
tatum by the presence of awned spikelets. The lat-
ter is distinguished by its annual habit, spikelets
1.8-2.2 mm long, upper glume with an awn 6-12.2
mm long. and lower lemma with awn 0.3—
long.
mm
Representative specimens. BRAZIL. Goiás: Niquelân-
dia, — Say nn india, Macedo, фе аурей Lope
2409 . Fonseca et al. 224 (MO), Oliveira et al. 61
623. К on 758 (IBGE. S1).
w
Paspalum buchtienii Hackel, Кереп. Spec.
153. 1908. TYPE: Bolivia.
La Paz: Prov. Sud Yungas, Sirupaya a Yana-
cachi, 2100 m, 1906, O. Buchtien 420 (holo-
type, W! photo, K*: isotype, US-2854141! pho-
o, SI!). Figures 8, €
Nov. Regni Veg. 6:
rhi-
zomes; floriferous culms 60-100 cm tall, ascendent.
Rhizomatous perennial, with leafy, long
erect or clambering, rigid, branching at the basal
and middle nodes; internodes 5-25 em long, 0.2—
0.3 cm diam., finely striate, glabrous, hollow, stra-
mineous or tinged with purple; nodes compressed,
brown, glabrous to shortly pilose. Sheaths 4-15 em
long, the lower ones overlapping, the upper ones
shorter than the internodes, glabrous or pilose, the
margins membranous, glabrous or ciliate with tu-
berculate hairs. Ligules membranous, 0.8—1.6 mm
long. brown, glabrous: pseudoligule absent. Blades
lanceolate, 0.6-1.7 flat-
tened, papyraceous, acute at the apex, glabrous to
6-16 cm long. cm wide.
shortly pilose on both surfaces, the margins gla-
brous or ciliate, shortly pseudopetiolate, pseudo-
petiole ciliate. Peduncles up to 28 cm long. terete,
glabrous. Terminal and axillary inflorescences pres-
ent, exserted, 10-15 em long, 6-9 cm wide. trun-
cate;
main axis 4-13 em long, flattened, glabrous:
racemes (4)7 to 11, alternate to subopposite. ar-
cuate, the basal ones 6-12 cm long, rachis of the
racemes flattened, 0.8—1.2(—1.4) mm wide, keeled
and with longitudinal nerves, green to purple. &
a-
brous. the margins scaberulous, ending in a spike-
let or in a naked point; pedicels paired, flattened
to triquetrous, glabrous, unequal. the upper up to
0.8 mm long; spikelets paired, occasionally the
lower one aborted, sparsely arranged in 2 to 4 se-
ries. Spikelets long-ellipsoid, 3-3.6 mm long, 0.8—
1(—1.3) mm wide, dorsiventrally compressed, acu-
minate, pilose, pale or tinged with purple. Upper
glume hyaline to membranous, 3-nerved, the mar-
gins corky, papillose-pilose with radiate, white
hairs, otherwise sparsely pilose. Lower lemma
membranous, 3-nerved.
elumiform, glabrous or
shortly and sparsely pilose on the surface, sca-
brous, with a depression at the base of the dorsal
surface. Upper anthecium long-ellipsoid, 2.2—
mm long, 0.8 mm wide, shorter than the spikelet.
plano-convex, membranous, stramineous, finely pa-
pillose with simple papillae over the surface, sca-
berulous toward the apex: upper lemma 5-nerved.
the margins flat and not covering the apex of the
palea. with silica bodies, microhairs, and hooks at
the apex: upper palea with microhairs at the apex:
lodicules 2. conduplicate, 0.3 mm long: stamens 3,
anthers 1.6 mm long; stigma plumose, lateral. Cary-
opsis nol seen.
Distribution and habitat. Paspalum buchtienti
grows in Bolivia (La Paz and Santa Cruz) and Peru
(see observations), where it is found on mountain
slopes on rocky or limestone soils along forest edg-
160 to 2100 m.
Paspalum buchtienii is closely related to P. hum-
es from
boldtianum, a species that is distinguished by its
herbaceous leaves, 2 to 6 ascendent. not arcuate.
racemes per inflorescence, each raceme (3—)5—9 ст
the
mm wide, and the spikelets, 1.2-1.4 mm wide, are
long: rachis of the racemes is usually 1.2-1.5
imbricate along the racemes.
It was not possible to examine the specimen Fos-
ter & D'Achille 12021 of Parque Nacional Manú,
which was cited by Tovar (1993) in his treatment
of the grasses of Peru.
The Steinbach 3015, Beck 8714,
17851. and Buchtien s.n. (К. SL, W) differ by hav-
ing just 4 to 5 racemes per inflorescence and nar-
specimens
row spikelets, sparsely arranged on the racemes.
Representative specimens. BOLIVIA, La Paz: Yungas.
Bang 265 (BM, К. MO, W); Prov. Nor Yungas. Coroico 10
km hacia Coripata, Beck 8714 (K, MO, SI, US): Prov. Nord
Yungas. cerca de 10 km al E de Coroico bai and a Car-
anavi. Beck et al. 19948 (К. LPB. SD: Nord Yungas. Polo-
Polo bei Coroico, Buchtien 3629 (MO, US, W); Nord Yun-
358 Annals of the
Missouri Botanical Garden
Figure 8. m buc к —A. Habit. —B. Detail of iue. — Portion of the rachis. —D. Spikelet, dorsal
— —E. 8 et, venti al v —F. Upper anthecium, dorsal view. —G. Upper antheci dud ventral view.
ryopsis, sc а —]. Ca aryopsis, hilum view. (A, B based on Mi el al. 19948, Sl; C-1 based on Ной &
mee 712, US.)
—
=
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
А А Р carinatum
* Р buchtienii |
| | |
\ 4 = 1
Y \ |
| | — |
\ \ |
200 400 600 800 1000 km | 1
ې ىى
AAA AA
© 100 200 300 400 500 600 miles
Prepared by Hendrik R. Ryphema
ah AS T — dL
90 40
Figure 9. Distribution of Paspalum buchtienii and P.
gas. Buchtien ong (BAA, МО);
Buchtien 447 (B
(К. SI. W). сап Cruz: Prov. A. Ibanez. Quebrada Sec:
road to Camiri, Nee 26848 (MO); Del Para, cerro del
boró, Steinbach 3015 (SI).
4. Paspalum burmanii Filgueiras, Morrone &
Zuloaga, Novon 11: 36, fig. 1. 2001. TYPE:
Brazil. Goiás: Mun. Niquelándia, 14725'5,
48°26' W, na da estrada entre Niquelán-
dia e Macedo, ca. 10 km N Niquelándia, es-
trada para — 20 May 1993, T. 5. Fil-
gueiras 2471 (holotype, IBGE! isotypes. MO!,
SI!, SP!, US!). Figure 10.
Rhizomatous perennials, cataphylls pilose: florif-
erous culms 1.5 m tall, erect, simple: internodes
12-17 em long, terete, hollow, lignified, glabrous,
smooth; nodes compressed, pilose, brown, Sheaths
15-22 cm long, longer than the internodes, papil-
lose-pilose, with the adaxial surface shiny, the mar-
gins shortly pilose. Ligules 2 mm long, membra-
Polo Polo bei Coroico,
BM, G, SD: € Yungas. Buchtien s.n.
> carinatum.
nous, brown; pseudoligule absent, collar pilose.
Blades lanceolate, 30-35 em long, 1-1.2 cm wide,
rounded, the apex acuminate, mostly basal, the dis-
al ones small, flat or with involute margins, the
basal ones papillose-pilose, midnerve conspicuous.
Peduncle subincluded, cylindrical, sparsely pilose
Inflorescences terminal, 20-25 cm long. 5-12 cm
wide; main axis 15-20 em long, glabrous: racemes
9 to 11, ascendent and divergent, alternate, the
basal ones 7-9 em long; pulvini densely pilose with
short and long hairs; rachis of the racemes 2.7-3
mm wide, flat, foliaceous, ending in a naked point,
occasionally in a spikelet, glabrous, green, the mid-
nerve Conspicuous and with anastomosed veins, the
margins scabrous: pedicels unequal, terete, sca-
brous; spikelets paired, imbricate and arranged in
. series. Spikelets ellipsoid, 3.3—4 mm long, 1-1.2
mm wide, pilose. Upper glume 3-nerved, acumi-
nate, membranous, pilose, more densely so toward
the base. the margins corky, hairy, with spreading
hairs up to 2 mm long. Lower lemma as long as the
A
360
Annals of the
Missouri Botanical Garden
P. ceresia
m P cymbiforme
e P cordatum
ж Р burmanii
А Р cachimboense
+ Р. goyasense
— — —
i-e
20) \ \ \
\ \
\ |
\ |
\ \
\
\
à 200 400 600 800 1000 km
о Ra 200 300 400 - Ын miles \
d by Hendrik K \ \
— жане A MRNA динанын... анаа m эрин
Figure 10. Distribution of Paspalum burmanii, P. cachimboense, P. ceresia, P. cordatum, P. cymbiforme, and P.
goyasense.
upper glume, 3-5-
0.8 mm wide, plano-convex, papyraceous, shiny,
with simple papillae all over the surface; upper
lemma 5-nerved, with silica bodies and prickles on
the upper portion, and bicellular microhairs and
macrohairs all over the surface; lodicules 2, 0.2 mm
long, conduplicate, hyaline; stamens 3, anthers 2
mm long; stigma plumose. Caryopsis obovoid,
mm long, 0.7 mm wide, brown; hilum punctiform,
basal; embryo less than Y the length of the cary-
opsis.
Distribution and habitat. Endemic of Nique-
lándia, in Goiás, central Brazil, on serpentine soils
Paspalum burmanii is related to P. niquelandiae,
which can be separated by its spikelets 2.3—3.3 mm
long, with the upper glume almost glabrous or
shortly hispid on the surface, racemes 8 to 30, ra-
chis of the racemes up to 1.5 mm wide, and pseu-
nerved, acuminate, membranous,
scabrous. Upper anthecium ellipsoid, 3 mm long,
doligule a ring of hairs 5-8 mm long. It is also
related to P. humboldtianum, which differs by its
plants 30-100 em tall, culms branching at the low-
er and middle nodes, otherwise unbranched, blades
5-21 em long, 0.5-1.8 em wide, evenly distributed
along the culms, and inflorescences smaller with 2
to 6 racemes, the rachis of the racemes 1.2-1.5 mm
wide.
Representative specimen. BRAZIL.
Goiás: Niquelán-
dia, a 10 km, Filgueiras 2477 (SI).
э. Paspalum cachimboense Davidse,
Morrone
& Zuloaga, Novon 11: 389, fig. 1. 2001.
TYPE: Brazil. Mato Grosso: Mun. Colider, es-
trada Santarém-Cuiabá, BR-163, km 762,
Serra do Cachimbo a 30 km de Guaranta,
92355, 54°55'W, 19 Apr. 1983, M. N. Silva,
I. L. Amaral, J. Lima, O. P. Monteiro & J.
Coélho 24 (holotype, MO!). Figure 10.
Volume 89, Number 3 Denham et a 361
2002 Paspalum subg. Ceresia
Caespitose annual: floriferous culms 20-70 cm кин — Hochst. ex Steud.. PI. Glumac.
tall, 1.5 em diam., red-tinged, geniculate, branch- 1855 [1953]. — carinatum Humb. А
ing at the middle and upper nodes; internodes hol- И) a bes fs o^ 2) m E ~ —
öll, ın Mart., А ras ). y
к r x 7 5 ‹ А y ho: a € 2 Кек *
low. glabrous: nodes brown, glabrous. Sheaths 3-8 rinam. Without locality, E ji Hostmann 1 306
em long, usually shorter than the internodes, dense- (holoi; ype, Pt isotypes, B!, BM! ! photo, SIL
ly pilose toward the distal portion, otherwise gla-
brous, the margins membranous. Ligules membra-
nous, 2—3 mm long; pseudoligule absent: collar
papillose-pilose. Blades linear to linear-lanceolate.
14-19 em long. 0.4-0.6 cm wide, attenuate at the
base, the apex acute, densely papillose-pilose on
both surfaces and along the margins, midnerve
manifest. Peduncles exserted, cylindrical. glabrous.
Terminal and axillary inflorescences present, 5-8
em long, 2-5 em wide: main axis absent: terminal
inflorescences with | to 4 approximate racemes, ax-
Шагу inflorescences usually with a single raceme:
pulvini pilose, usually with a membranous bract гир
to 5 mm long: rachis of the racemes winged,
Y
—§
9
em long, 2-3 mm wide, ending in a naked —
glabrous, the midnerve manifest and the margins
membranous. nerveless; pedicels flat. glabrous:
spikelets solitary, imbricate and arranged in 2 se-
ries. Spikelets long-ellipsoid, 4—4.3 mm long. l-
1.1 mm wide, pale, pilose, occasionally glabrous.
Upper glume membranous, rounded at the base. the
apex acute, 3-nerved, with one central and two sub-
marginal nerves, pilose at the base, the apex shortly
ciliate. the rest of the surface glabrous. Lower lem-
ma glumiform, as long as the upper glume, rounded
at the base, 3-nerved, the margins with mixed pa-
pillose and not papillose hairs. the apex shortly cil-
iate. Upper anthecium ellipsoid, 2.2-2.5 mm long.
0.8 mm wide, shorter than the spikelet, stipitate.
membranous; upper lemma with simple papillae.
macrohairs, and bicellular microhairs toward the
distal portion: lodicules 2. conduplicate. hyaline:
stamens 3. anthers 1.5 mm long: stigmas 2. plu-
тозе. Caryopsis not seen,
Distribution and habitat. Endemic to Serra do
Cachimbo in central Brazil, where it is known only
from the type collection.
Paspalum cachimboense is a member of section
Pectinata and closely related to P. lanciflorum,
which differs by its perennial habit, with simple
and erect culms. rachis of the racemes (4—)5—7 mm
wide, spikelets (4.8-)5.6-7.2 mm long. and upper
anthecium 3.2—1.7 mm long.
6
Paspalum carinatum Humboldt & Bonpland
ex Flüggé. Gram. Monogr., Paspalum: 65.
1810 [as “Paspalum ipsi TYPE: Co-
lombia. Without locality. Н. A. von Hum-
boldt & A. J. A. am ae s.n. (holotype, B-W
not seen: isotypes. BM! photo. SI. US-
2942176!). Figure 9.
O0-102047!. P!, US- Aen т
де» spissum = Swale "n. е 14: 358. 1967. уп.
TYPE: Brazil. Магапһао:
tonio de Balsas, 20-25 Mar.
4050 (holotype, US-1612651! photo. SID.
Carolina to San An-
934 R. Swallen
xo
Short-rhizomatous perennial; floriferous culms
(25-)40-70 ст long, finely striate, glabrous: nodes
brown, glabrous to shortly pilose. Sheaths 5-13 cm
long. usually shorter than the internodes, glabrous
to hirsute, the margins membranous. Ligules mem-
branous, ca. 0.6 mm long, brown, glabrous: pseu-
doligule absent or present, when present a ring of
white hairs. Blades filiform, 5-20 cm long, 0.5-1(—
3) mm wide. involute, erect, mostly basal. rounded
at the base, the apex setaceous, papillose-pilose on
both surfaces, more densely so toward the base of
the adaxial surface, the margins involute, ciliate.
Peduncles up to 35 em long. subterete. glabrous.
Terminal inflorescences 6-12 em long, with a mem-
branous bract often at the basal node: main axis
absent or up to 3 em long, flat, glabrous: pulvini
glabrous: racemes 1(ог 2). alternate when 2 ra-
cemes are present, straight or subfalcate. 5-12 ст
1.8-2.5(—3.5)
ending in a naked point, glabrous, the
long: rachis of the racemes winged.
mm wide,
midnerve conspicuous, the margins membranous to
hyaline, partially covering the spikelets: pedicels
short. flat: spikelets solitary, densely imbricate in 2
series. Spikelets long-ellipsoid. 3.3-5 mm long. 1—
dorsiventrally compressed. pale.
1.2 mm wide.
densely pilose at the basal portion. the apex acute
r rounded. Upper glume as long as the spikelet.
membranous, 3-nerved, densely pilose on the lower
portion. with white hairs up to 3 mm long. shortly
pilose on the apex, otherwise scabrous. Lower lem-
ma glumiform. slightly shorter than the upper
glume, 3-nerved, pilose on the lower portion, the
hairs shorter than those of the upper glume, shortly
pilose and scabrous at the apex, the rest of the
surface glabrous. Upper anthecium long-ellipsoid.
2.7-3.6 mm long. | mm wide, 0.6-1.8 mm shorter
than the upper glume, plano-convex, flattened to-
ward the distal portion, membranous, shiny, finely
papillose with simple papillae all over the surface;
upper lemma with macrohairs, bicellular micro-
hairs, and prickles at the upper portion; upper pa-
lea with bicellular microhairs and macrohairs at the
lodicules 2. 0.2
apex: mm long. conduplicate, hy-
362
Annals of the
Missouri Botanical Garden
aline; stamens 3, anthers 1.6—1.8 mm long; stigmas
2, plumose, of lateral emergence. Caryopsis ovoid,
1.6 mm long, 0.8 mm wide; hilum elliptic, embryo
less than half the length of the caryopsis.
Iconography. Smith et al.. Fl. Il. Catarinense:
918, fig. e-g. 1982. Renvoize, The Grasses of Ba-
hia: 214, fig. 79, A. 1984. Judziewicz, Fl. Guianas:
469, fig. 83, D. 1990. Rodríguez, Ernstia 8(2-3):
37. 1998
Common names. oasis ciel (Brazil);
Pa-mac” (Guyana).
Chromosome number. = 10 (Davidse & Pohl,
1978); n = 40 (Davidse & Pohl, 1972).
Distribution and habitat. Trinidad and Tobago,
"Garcita" (Venezuela); *
Nicaragua, Colombia, Venezuela, Guyana, Surinam,
Brazil, and Bolivia. It grows in humid savannas on
sandy, rocky, or limestone soils, in open places with
periodical fires, between 100 and 1600 m.
Paspalum carinatum is related to Р. goyasense
and P. stellatum. Paspalum goyasense has inflores-
cences with two distant racemes, rachis of the ra-
cemes 1-1.6 mm wide, ending in a developed
spikelet, spikelets with the upper glume densely
pilose in the lower half, otherwise glabrous, upper
glume and lower lemma slightly convex, and blades
linear, 8-20 cm long, 0.2-0.4 ст wide. Paspalum
stellatum has rachis of the racemes wider [(4—)5—
10 mm]; when two racemes are present, they are
conjugate, upper glume and lower lemma with long
ciliate margins, with hairs up to 4 mm long, and
the upper anthecium is 0.4—1 mm shorter than the
upper glume and lower lemma.
A rudimentary membranous bract is common in
the basal portion of the racemes in Paspalum car-
inatum, as was pointed out by Sendulsky and Bur-
man (1978).
Swallen (1967) related P spissum with P. cari-
natum, characterizing the former by its densely pa-
pillose-pilose blades 2-3 mm wide, and appressed
to the culms. After examination of many specimens
and the type collections, we conclude that there is
a notorious variation in these characters: filiform
blades are mixed with linear blades in several
specimens, such as Valls & Silva 8340, Irwin et al.
27667, and Swallen 3687, 4078; others have fili-
form blades with papillose-pilose hairs. e.g., Dav-
idse & Huber 15436 and Blydenstein & Saravia
1082. Therefore, P. spissum is here considered a
synonym of P. carinatum.
Although blades are predominantly basal, cau-
line blades were present in several specimens. This
is probably due to the absence of periodical fires,
which are associated with its savanna habitat.
Representative specimens. BOLIVIA. Beni: Prov. Vaca
Diez, Riberalta, Beck 20564 (K). Santa Cruz: Prov. Ve-
lasco, San Ignacio, Bruderreck 2 (K, LPB, SI); Velasco, 5
km E of Comunidad Carmen del Ruiz, Killeen 2823 (LPB,
MO, US). BRAZIL. Amapá: Macapá, Black & Froés 51—
12291 ips Mun. Calcoene, Mori & Cardoso 17298 (MO).
Bahia: 15 km S of intersection of ue BR-020 and Rio
Roda Ve lha, Davidse et al. 12 M MO); 51 km E of Bar-
reiras along Hwy. BR-242, ia et al. 12124 (MO).
Distrito Federal: Brazilandia, Allem & Vieira 1558
0); km from Brasília on — road, Clayton
4804 (K, US). Goiás: Serra pc ca. 15 km S of Goiás
Velho, Anderson — (COL, К, MO); Barra da Lagoa For-
mosa, Glaziou 22547 (G, K, P, W); Formosa, Glaziou
22556 (К, Р, US, W). Maranhão: Barra do Corda to Gra-
jahú, d 3643 (Р); Barra do Corda to Grajahú, Swal-
pie 3 (US); Carolina a San Antonio de Balsas, Sw а
— Mato Grosso: Mun. Amambaf, 12
бу -Iguatemi, Allem & Vieira 2007 (МО); са. "270
km N of Xavantina, Ramos & Sousa 182a (K, US). Minas
Gerais: Metallurgica, Serra de Ouro Branco, Chase 10306
S): Serra do Espinhaço, 8 km E de Diamantina, /rwin
et al. 27667 (MO). Para: Belterra, beira do Lago Jurucuf,
—
—
. Ponta Grossa,
(К, P, SI, US). Rio Grande do Norte: Estremoz to Natal,
Swallen 4762 (US). Roraima: along Boa Vista—Bonfim
road (BR-401), km 40, Coradin & — 696 (MO).
Santa Catarina: Rio Capinsel, Dusé 3 IS).
Sao Paulo: Casa Branca, — 10595 (MO, US). To-
cantins: Пћа do Bananal, Mun. Lagoa da — Re-
serva indígena — da Ely et al. 4082 (IBGE, S
COLOMBIA. Arauca: Estación de Cravo Norte, Llano 4
(US). Boyacá: | llanos Orientales, al sur de El Yopal, Bh-
denstein & Saravia 1082 (COL). Cundinamarca: savan-
na of San Martín, 100 mi. SE of Bogotá, Shaw s.n. (US-
1343796). Guaviare: Mun. San José del Guaviare,
serranía La Lindosa, Ciraldo.Cafias & López 2566 (MO).
eta: Llanos Orientales, entrada de la Serranía San Mar-
PE
). Vichada: — ls El Tuparro, entre
, El Tapón, Daniel 60 (COL).
. Weruni- luni. savanna, Abraham 83 (К, US);
Naroni District, Lethem, 4 mi. N of — >m, Irwin 739
(US). NICARAGUA. Zelaya: ca. 5 km S of Bilwaskarma
on road to Puerto Cabezas. Pohl & — 12274 (MO).
TRINIDAD AND TOBAGO. Erin savanna, S of Point For-
tin, Soderstrom 1124 (US). VENEZUELA. Amazonas: 3—
6 km N de S sien sobre ruta a Puerto Ayacucho, Dav-
idse et al. 16775 ( Anzoátegui: Distr. Bolívar, just
S of El а Fila El Purgatorio, Davidse & Gonzdlez
19317 (MO); Distr. Freites, Burro trail betw. San Durrial
& Los Pajaritos, Davidse & Gonzdlez 19802 (MO). Apure:
S of San Fernando de Apure. Alba 53-96 (US). Bolivar:
Dpto. Heres, margen izquierda de la Quebrada Aguas Ne-
gra, 10 km al E del tepuy El Zumbador, Aynard 6019
(MO). Guárico: Estación Biológica Los E anos, del M. |
C. Calabozo, Aristeguieta 4219 — 05
NWN of San Felix & Rio Orinoco along Hess
turín, Davidse et al. 4615a (MO). Zulia: DE. — 9
km S of the Machiques intersection along the Maracaibo-
La Fria Hwy. (Hwy. 6), Davidse et al. 18355 (K )
los ríos Tomo y El Tupa
ANA
Volume 89, Number 3
2002
Denham et al. 363
Paspalum subg. Ceresia
7. Paspalum ceresia (Kuntze) Chase, Contr. U.S
Natl. Herb. 24: 153. 1925. Paspalum mem-
branaceum Lam., Tabl. Eneyel. 1: 1791,
hom. Шер. Ceresia elegans Pers., Syn. Pl. 1:
85. 1805. Ceresia membranacea wd P.
Beauv., Ess. Agrostogr.: 9 y 171, t. 5. f. 4.
1812. Paspalum elegans (Pers.) — &
Schult., Syst. Veg. 2: 290. 1817, comb. Шер.
Panicum ceresia Kuntze, Revis. Gen. Pl. 3:
360. 1898. TYPE: Peru. Without locality.
without collector (holotype, P-LA! photo, SI;
S-2855815!). Figures 10. |
isotypes, Р!, US
Каре — eum Lam. var. pir iia Doll.
‚ Fl. Bras. 2(2): 94. 1877. TYPE: Bolivia. La
Par айй, 2600-2700 т. Feb.-Apr. 1861. б. Man-
don 1255 — designated here, №1: isolecto-
types, BM!, СІ, КІ, P!
— dona м ин Lain var. не Doll. i
2: 94. 1877. TYPE: Brazil.
campos between Natividade and aoe eição,
3. Gardner 4029 (lectotype, designated here
В!; isolectotypes, BM!, G!, K!, PL. US-80085!. WD).
..
lart.. . bras. Golds:
‘eb.
D d
—
м
~
Caespitose, short rhizomatous perennial; rhi-
zomes arcuate, covered by pilose cataphylls: florif-
10-80 cm tall, 0.1-0.2 cm diam.,
branched or unbranched; internodes 4—12 cm long.
erous culms
terete, hollow, glabrous; nodes glabrous. Sheaths 3—
12 cm long, glaucous, glabrous to sparsely pilose
on the distal portion, the margins membranous, gla-
brous. trun-
cate, glabrous: pseudoligule absent, collar glabrous.
Ligules 1-2 mm long, membranous,
Blades filiform, linear-lanceolate or lanceolate, 7—
21 em long, 0.3-1.5(-2) em wide. flat or with in-
volute margins, rigid, slightly divergent, glaucous,
glabrous to sparsely papillose-pilose toward the
base of the adaxial surface, the margins glabrous
to sparsely papillose-pilose. Peduncles subincluded
to exserted, up to 35 em long. cylindrical, glabrous.
Terminal inflorescences exserted, 4-12 cm long. 2—
10 em wide; main axis absent or up to 7 em long.
triquetrous, glabrous; pulvini glabrous or shortly pi-
lose, without papyraceous bracts; racemes (1)2 to
7. ascendent, divergent and alternate, distant: ra-
chis of the racemes 2-7 cm long, 6-10 mm wide.
ending in a naked point, glabrous, green or purple
at the middle portion, hyaline at the margins, these
brown or ferruginous; pedicels up to 0.4 mm long.
flat. hispidulous; spikelets solitary and imbricate in
2 series. Spikelets long-ellipsoid, (3—)3.1—3.0 mm
long, 1.2-1.4 mm wide, acute, villous, early decid-
uous: upper glume and lower lemma subequal. Up-
per glume hyaline to membranous, 3-nerved with
one central and two marginal nerves, densely pilose
at the base and along the margins, the hairs silky,
ascendent, up to 4 mm long. Lower lemma glumi-
form, hyaline, pilose, with 2 submarginal nerves.
Upper anthecium long-ovoid, 2.6-3.6 mm long. 1—
1.2
white, scabrous and pilose toward the apex: upper
mm wide, not stipitate, hyaline and shiny.
lemma with prickles, macrohairs, bicellular micro-
hairs, and silica bodies at the apical portion, upper
palea with bicellular microhairs and macrohairs at
the apex: lodicules absent; stamens 3, anthers 2.4—
2.8 mm long. Caryopsis obovoid, 1.2-1.6 mm long.
embryo Y the
0.8 mm wide; hilum punctiform.,
length of the caryopsis.
Distribution and habitat. Ecuador, Peru. Brazil,
Paraguay. Bolivia, and Argentina. in rocky moun-
tain slopes and open fields on sandy and rocky
soils. between 300 and 3000 m.
Paspalum ceresia is related to P. stellatum, a spe-
cies with filiform leaves 8-27 em long, 0.1—0.3 em
wide, 1 (occasionally 2) racemes, | or 2 papyra-
ceous bracts in the basal pulvini, radiate hairs on
the pedicels, lodicules present, and upper anthe-
cium stipitate. Paspalum ceresia is also similar to
P. trachycoleon, a species from Central America,
Venezuela, Colombia, and southern Brazil. which is
characterized by the rachis of the racemes 4—0 mm
wide, spikelets paired, with an annular thickening
al the 3-
nerved.
А single collection was examined from Paraguay,
base. and lower lemma membranous.
Isabell s.n. at Р, a specimen without any information
about collection date or locality; it is possible this
specimen. was mislabeled, considering the distri-
bution range of the species.
There is considerable variation in the leaf shape,
from filiform to linear and rigid, up to 0.3 cm wide.
in the specimens Oliveira et al. 305, 598, Pinheiro
& Carvalho 142. ‚ 625, Gardner 4029, Brooks
et al. 125, Eiten & Biten 4418. 3777, 10789, and
Swallen 3772, single
specimen, e.g., Souza et al. 5548, Filgueiras 2280,
as in
filiform and lanceolate in a
or with lanceolate leaves, up to 2 cm wide,
Doell (1877) described variety aequiglume and
variety inaequiglume of P. membranaceum on the
basis of variation of the upper glume and lower
lemma length, a variable character according the
degree of maturity of the spikelets. This author cit-
ed the specimens Rhodé s.n. and Mandon 1255 as
syntypes of variety aequiglume, of which Mandon
1255 agrees with the protologue and is here se-
lected as lectotype of this taxon. Also, Doell (1877)
cited Pohl s.n. and Gardner 4029 as syntypes of
variety inaequiglume, of which Gardner 4029 is
here chosen as lectotype of the variety.
364 Annals of the
Missouri Botanical Garden
Figure 1 l. Paspalum ceresia.—A. Habit. —B. Detail of ligule. С. — of the rachis. —D. Spikelet, —
view. — pikelet, ventral view. =. Upper anthecium, dorsal view. —G. Up pper anthecium, ventral view. —H. Upper
lemma. —1. C: агу ро, scutellar view. —J. Caryopsis, hilum view. (A-H based on Hunziker et al. 10354. Sl: 1, J based
on i et al. 22135, SL.)
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
ARGENTINA. ee re
Santa Bárbara, El F . Cabrera & Fabris 22730 (US
Dpto. Que ‘a, Sier 2 Zenta, Venturi 8320 4 LIL.
MO, P, SI, W). Salta pt Candelaria, Unquillo, Schrei-
ter deso "a IL): Dpto. ( Guac hipas, Alemania, Venturi 9905
( n Pedro de Colalao, Dinelli 12876
(1. Т ). BOLIVIA. € анча : 15 km S of Padilla, Ren-
voize & Cope 3870 (K, | Cochabamba: Prov. Quil-
ac 45 km А Oruro, Beck 4025 (US). Potosí:
Prov. Charcas, — L ama — Mostacedo 314 (MO
Santa Cruz: Prov. Caballero, 15 km arriba de San —
de Portero, Killeen ‘ iuh 4152 (MO); Prov. F lorida Es
Río Mew 3 km N of hwy. at Yerba ns na, Nee 403
(К, MO arija: d Fiebrig 2664 (SI). B RAZI.
Goias: ES . Buracáo, ca. 35 km (bs road) N of Goianésia,
Brooks et al. 125 (К, MO, NY): Niquelândia, ca. 18 km
W e Niquelándia, Filgueiras 2280 (MO, US): Niquelan-
dia, 2 km da e y , : de Chao para o Macedo Velho, Oliv-
Mun. Niquelándia, Macedo. ca. 0.
ша, Oliveira et al. 508 y S).
le Eales region, Kite n ¢
Representative —
W). Tucumán: 5
eira et al. 30:
km abaixo da mina de
Maranhao: Mun. —
K, P). iia Gerais Retiro ca. 16 km и»
{ 5548 (К
ara: Estreito-Maraba, km | " 19 ‚ре ro &
Carvalho 142 (US); km 2, Pinheiro & Со, 2 517, 025
ECUADOR. Azuay: valley of the Río Paute. be-
tween Paute and Cuenca, between Ríos Azogues & Gual-
aceo, Camp 2332 (K, P, US, W) soja: road San Pedro
de la Bendita (W of C O asne, Ollgaard et al.
90721 (MO): 6 km W of Colaisaca on road to Sozoranga.
Peterson & Judziewicz eon (К. MO, US). PARAGUAY
Without de ‘partment: without locality, Isabell s.n. (P).
PERU. Apurimac: Prov. Abancay, Cunyac, Chávez 3217
(MO). Cajamarea: Prov. kl Ti
la carretera. Coc nO
(GSL US); Dpto. Cajamarca, 4 km f Cele ndin on road to
оны & Wright 519 © )). Cuzeo:
Pep. unc Mollepata, Vargas 19 (M A). Huancaveli-
— 1020 (US).
ea: entre Izcuchaca y Acoria,
8. Paspalum cordatum Hackel, Ark. Bot. 9(15):
1910. TYPE: Brazil. Paraná:
der Nähe des Stüdtchens Рома Grossa, am
Rande eines Sumples, 7 Ene. 1904, P. K. H.
Dusén 3248 (holotype, W!: isotypes. Gl. US-
2942131!). Figures ТО, 12.
Rio Tibagy in
Robust, rhizomatous perennials with long. arcu-
100—130 cm
unbranched: internodes finely
ate rhizomes: floriferous culms erect,
tall, 0.4—1 em diam.,
glabrous; nodes compressed, gla-
striate, hollow,
brous, brown. Sheaths usually longer than the in-
ternodes, the lower ones overlapping. loose, densely
hirsute on the basal portion or glabrous. the inner
surface shiny. brown to. red-tinged, the margins
membranous. Ligules membranous, 1-1.3 mm long.
brown, glabrous: pseudoligule a ring of dense. rigid
hairs up to 6 mm long; collar densely pilose. Blades
linear, 10-50 cm long. 0.5-1.1 em wide, flat. the
margins involute, ascendent and rigid, densely vil-
lous on both surfaces, attenuated at the base. the
apex beaked. Peduncles exserted, up to 45 em long.
subterete, glabrous. Terminal inflorescences exsert-
ed. 7-18 cm long, 6-18 em wide, subdigitate: main
axis up to 4 em long, wavy, glabrous to sparsely
pilose; pulvini pilose; racemes 5 to 10, 12-16 em
long, ascendent, alternate to subopposite, ending in
a naked point; rachis of the racemes flat, 1.8-2.4
(-3.2) mm wide, glabrous and green, the margins
shortly winged, scaberulous; pedicels short. gla-
brous; spikelets solitary, densely imbricate in 2 se-
ries. Spikelets ovoid, 4—5(—6) mm long. 2.5—3.5 mm
wide, dorsiventrally compressed, pale, winged, cor-
date, the apex acute. Upper glume as long as the
spikelet, membranous, glabrous, cordate, 5-nerved,
with 3 central nerves reaching the apex, and 2 lat-
eral nerves only notorious at the base. Lower lemma
3.4—3.6(—5
nerved, the margins with rigid and thick papillose-
mm long, | mm wide, subcordate, 3-
—
pilose hairs, not winged, shortly pilose. Upper an-
thecium ellipsoid, membranous, 3-3.2 mm long.
0.8 mm wide, shorter than the spikelet, plano-con-
vex, pilose on the upper margins, pale: upper lem-
ma with simple papillae all over the surface. and
bicellular microhairs, macrohairs, and silica bodies
toward the apex: upper palea with simple papillae.
bicellular microhairs, and macrohairs on the upper
portion: lodicules 2, са. 0.4 mm long. conduplicate.
hyaline. glabrous: stamens 3, anthers 1.6 mm long:
lateral. Caryopsis not
styles 2. stigma plumose,
seen.
Distribution and habitat. This species grows in
southern Brazil. in the states of Amazonas. Paraná.
Goiás. Minas Gerais, and Sáo Paulo. also in Colom-
bia. in the states of Meta and Vichada. in open
areas, lowlands, and swamps.
Rodríguez (1998) considered Paspalum corda-
tum as a synonym of P pectinatum without indi-
cating the criteria for such a decision. Both species
share spikelets with the upper lemma not winged
and with conspicuous rigid, papillose-pilose hairs
on the margins. Nevertheless, we consider P. pec-
tinatum a different entity by including specimens
—8 cm
long. per inflorescence. Additionally, plants, which
with 2. occasionally 3, conjugate racemes,
grow in savannas or open fields on sandy or rocky
soils. are smaller and more densely caespitose than
those of P cordatum. with lower sheaths burnt by
fire.
Paspalum cordatum is also related to P. imbri-
catum, this species with glabrous spikelets and low-
er lemma winged.
Note that in the specimens Blydenstein 939 and
Cabrera 2278, the rachis of the racemes is 3—3.2
mm wide. with crenate margins, and spikelets are
bigger, 5.2-6 mm long.
eno
366 Annals of the
Missouri Botanical Garden
4 — ZZ
— — > ЕРІ FZ LIES OS
P".
IA
мер,
—
>
N
^ 29
—
Da
SR
Paspalum cordatum.—A. Habit. —B. Detail of ligule. —C. Portion of the rachis.
—G. Upper anthecium, ventral view. —
US
Figure 12. —D. Spikelet, —
view. —E. Spikelet, ventral view. —F. Upper anthecium, dorsal view. . Up
lemma, lodicules, and gynoecium. (Based on Hatschbach & Renvoize 5276, US.)
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
Representative specimens. BRAZIL. Amazonas:
ا Hwy.. 53 km W de Aripuana River, i
“ = 8 (MO, US- AR 34). Goiás: Jataí, cedo
T — Gerais: without locality, с 1699
"à 1 Turma, in paludosis, Dusén 9109 (BM.
US): н Dusén 16174 (BM, G, SI, US); Mun.
Campo — Serra S. Luis, Hatschbach et al. 12097 (К.
MBM): V а En Bul Serra * Fur rnas, Smith et al.
14579 (K. PS ; Mun. - Rio
Funil. Smith et Py 11858 (К.
hne Brade 19638 (US). Sio Paulo: Estação Flo-
tal de Paraguaqu Paulista, 6 km N of city, Clayton 4651
(К. US). COLOMBIA. Meta: Llanos orientales, Hato Po-
zones, Blydenstein 939 M ~ MO). Vichada: Caño Ur-
imica, Cabrera 2278 (Cl
9. Paspalum еа E. Fournier, Mexic.
m. 2:5. TYPE: Mexico. México: San
Pablo, А "s jose. 226 (lectotype. desig-
nated by Chase (1929: 20), C not seen). Figure
Shortly rhizomatous perennial; floriferous culms
60-90 cm tall, branching at the middle nodes or
simple: internodes terete, striate. glabrous: nodes
brown, covered by ascendent white hairs. Sheaths
longer than the internodes. the
margins ciliate at the lower portion. Ligules brown.
membranous, 1-1.2 mm long: pseudoligule a ring
of hairs. Blades linear, 5-19(-24) cm long. 0.2—
0.5(-0.8) em wide, flat, glabrous or shortly pilose
on the adaxial surface. the margins with tuberculate
striate, glabrous,
—
hairs up to 5 mm long, shortly pseudopetiolate,
pseudopetiole with short, white hairs. Peduncles
exserted, Terminal
inflorescences 8—12 cm long, 3—4 cm wide:
axis absent or up to 5 em long, shortly hispid or
3-8 cm
long. alternate. ending in a naked point: rachis of
the racemes 2-2.8 mm wide, green and tinged with
purple. the midnerve manifest, wings short, hya-
15-20 cm long, terete, pilose.
main
glabrous: pulvini pilose; racemes 1 to 4,
line, nerveless: pedicels ca. 0.3 mm long. pubes-
cent, spikelets paired and imbricate in 2—4 series,
usually the lower spikelet aborted. Spikelets long-
| mm wide, acute. pilose.
ellipsoid, 3 mm long.
pale. Upper glume hyaline to membranous, 3-
nerved, with one central and two submarginal con-
spicuous nerves, the margins corky, with long, +
ascendent hairs. up to 3 mm long. and with small
hairs on the basal portion. Lower lemma as long as
the upper glume, membranous. 3-nerved, shortly
pilose toward the apex. otherwise glabrous, de-
pressed on the basal portion. Upper anthecium
long-ellipsoid, 2.2 mm long. 0.8 mm wide,
membranous: upper lemma 3-nerved, finely papil-
pale.
lose and with prickles, silica bodies, and bicellular
microhairs at the apex: upper palea papillose. with
macrohairs and bicellular microhairs at the apex:
lodicules 2, 0.2-0.3 mm long, conduplicate, hya-
line: stamens 3, anthers 1.5 mm long: stigma plu-
mose, lateral. Caryopsis not seen.
Iconography. Chase, Contr. U.S. Natl. Herb.
28(1): 21, fig. 4. 1929
Distribution and habitat.
Mexico to Nicaragua, on mountain slopes on rocky
soils. between 750 and 1300 m.
This species differs from P. humboldtianum and
P. polyphyllum by the rachis 2-2.8 mm wide (vs.
1.2-1.5 mm in Р. humboldtianum and 1-1.2 mm
in P. polyphyllum), and by the presence of hairs at
the apical portion of the lower lemma.
The specimen Davidse et al. 31637 has 2 to 7
Found from southern
racemes per inflorescence.
еги — EL SALVADOR. La Lib-
ertad: Eggler 639 (US). Santa Ana: Mun. Candelaria de
la Frontera, — El Vupe, Linares 3663 (MEXU). GUA-
MALA. Zacapa: lower slopes of Sierra de las Minas,
along trail above Río Hondo. Steyermark 29550 (US).
HONDURAS. Distrito Federal: Llamapa, SW of Talan-
ga. y um Santa Regina de Archaga. Pohl & Gabel
13531 (MO). Francisco Morazán: 20 km N of Pespire
"pie Hwy. | to Tegucigalpa, Davidse & Pilz 31637 (MO):
Mesas, Swallen 10809 — US). 10747
S). Olancho: El Espino, 8 km NE de Catacamas, En-
— 23 (MO). MEXICO. Consoquitla. Liebmann 2
(MO): Frontera dg 6-8 km E
C cea aa pose road to Ciudad Cuauhtemoc,
39106 (MEXU). Chiapas: "Mir xtapa, along Mexican
Hwy. 190 in the т атап Paraje of Muctajoc, Breedlove
13822 (MEXU,
ay bake 532 (BA. * b. NICARAGUA. Chinandega: Vol-
sita. Montaña del Cielo, 12742' N. 06758 W. Rueda
e d 1200 (MO).
10. dene eucomum Nees ex Trinius, Sp.
Gram. l: pl. 110. 1828. Pas c eucomum
Nees, "Y ie Enum. РІ. . hom.
illeg. TYPE: Brazil. W — me Ай» r ‘alo
isotypes. BM!,
US-285 1605).
s.n. (holotype, LE not seen:
photo, SI!. P!, US-2854694!,
Figures 13. 14.
Paspalum — Nees ex Trin. var. — Doll,
Mart.. Fl. Bras. 2(2): 65. 1877. TYPE: Brazil. Minas
Gerais: Lago Santa, Ё. н аа s.n. jest nol
ated; isotype, US-28546%
Paspalum plende ns Hack.. oin n Z. 51: 238. 1901.
azil. Goiás: without locality, A. КОМ. Gla-
ziou p 55 51 (lectotype, designated here, W!).
Caespitose, shortly — I florif-
erous culms 60-110 cm tall, —0.3 cm diam..
erect, ascendent, unbranched, — branch-
ing at the middle nodes: internodes 6-15 cm long.
nodes glabrous. Sheaths usually
striate, glabrous:
longer than the internodes, 8-25 em long. overlap-
ping. glabrous or papillose-pilose all over the sur-
Annals of the
368
Missouri Botanical Garden
1mm
— — = — © == == =
— ===> A ===
RI SS — NS
WA | INS
— 9
1mm
y
M SSS
— —
Portion of the rachis. —D. Spikelet, dorsal
abit.
B. Detail of ligule. —C.
Paspalum eucomum.—A. H
Figure 13.
». Upper anthecium, ventral view. —H. Upper
view,
E. Spikelet, ventral view. —F. Upper anthecium, dorsal view. —(
palea, lodicules, and anthers. (Based on Chase 9097, US.)
Volume 89, Number 3
2002
Denham e
al. 369
Paspalum ds Ceresia
€ P. humboldtianum
ж P reticulinerve
W P eucomum
| \
o 200 400 600 800 1000 km
AA) \
— —
0 UM 200 300 400 0 500 eno miles \ 7
repared by Hendrik R ma \ i
— — — —— —
100 36
Г 70
— —
L— — ——————— Е —
Figure 14.
face or only at the distal portion, the hairs up to 5
mm long,
with one margin ciliate. Ligules membranous, 0.2—
0.5 mm long, truncate, brown, glabrous; pseudoli-
gule a ring of white hairs up to 2.5 mm long. Blades
10-30 cm long, 0.2-0.5 em wide,
flexuous. rigid and involute, narrow, the apex se-
filiform, erect or
taceous, adaxial surface with conspicuous papillae
and tuberculate hairs, abaxial surface glabrous or
both surfaces densely hirsute. Peduncles exserted,
up to 30 ст long, subterete, pale or purple, gla-
brous. Terminal inflorescences conjugate; main axis
absent; pulvini pilose, with a bract usually present:
racemes
2. (4-)6-17 cm long. occasionally with a
third approximate raceme, ascendent, divergent,
ending in a naked point: rachis of the racemes 1.5—
2.54) mm wide, shortly winged, glabrous, green
to purple, the margins hyaline, nerveless; pedicels
short, flat, and sparsely pilose; spikelets solitary,
densely imbricate in 2 series. Spikelets ellipsoid,
the margins membranous, glabrous or
Distribution of Paspalum eucomum, P. humboldtianum, and P. reticulinerve.
2.5-3 mm long. 1.4-1.6 mm wide, dorsiventrally
compressed, acute, densely pilose toward the base
and the margins. pale or tinged with purple. U
glume as long as the spikelet,
nerved, the
pper
membranous, 3-
base and margins covered with white,
rigid hairs, the rest of the surface with short and
appressed hairs. Lower lemma glumiform, slightly
shorter than the upper glume, 3-nerved. Upper an-
thecium obovoid. 2.4— mm long, 1.2 mm wide,
membranous, pale, shiny, rounded at the apex.
smooth: lodicules 2, ca. 0.3 mm long, membranous,
conduplicate; stamens 3, anthers 1.5 mm long; stig-
ma plumose, lateral. Caryopsis not seen.
Distribution and habitat. Restricted to central
and southern Brazil in the states of Goiás, Minas
Gerais, São Paulo, Paraná, and in the Federal Dis-
trict. It grows in cerrados, in open fields on sandy
between 700 and 1100 m.
Paspalum eucomum is closely related to P. mal-
soils,
370 Annals of the
Missouri Botanical Garden
meanum. The latter species mainly differs by the 20 em long, 0.1 cm diam., terete, hollow, glabrous:
spikelets 1.6-2 mm (vs. 2.5-3 mm long): also, P.
malmeanum grows in Bolivia and Mato Grosso, on
flooded soils, while P. eucomum occurs in central
and southern Brazil, usually in drier settings. lt
should be pointed out that spikelet size, 2.3-2.5
mm long, is intermediate between both species in
Macedo 351, Sendulsky 1239, and Glaziou 17406.
These species are here considered as different en-
tities until more material becomes available.
Chase (ined.) distinguished P. eucomum from P.
splendens by the spikelet size (2.5 mm long vs. :
mm), rachis width (1.5-2 mm vs. 2-3.5 mm), and
general aspect of plants (delicate vs. robust). Upon
—
study of many specimens, we conclude that there
is a gradation in these characters, which are not
useful to distinguish both species.
Chase (ined.) and Sendulsky and Burman (1980)
pointed out that Glaziou 22555 is one of the syn-
types of Paspalum splendens. Nevertheless, Hackel
(1901) cited, as syntypes, the specimens Glaziou
22552, 22553, and 22554, of which Glaziou 22554
agrees with the original diagnosis and is here des-
ignated as lectotype of the species.
The rachis of the racemes is exceptionally wide,
up to 4 mm, in the specimens Dias et al. 67 and
Weddell 2556.
Chase (ined.) studied, at B, two specimens col-
lected by Sellow, 1238 and 1336,
amined by Nees and probably represent isotypes of
which were ex-
P. eucomum. Duplicates of these collections from
US were identified by us as P. eucomum.
Representative specimens. BRAZIL. Distrito Fed-
eral: Brasília, de Jesus 3 (US); Brasflia, Dias et al. 67
E : re of Lagoa Paranoá, Irwin et al. 15328 (K,
MO, US). Goias: without н lity, Glaziou 22551 (B, G.
К, E. Quo 22555 5). Minas Gerais: be-
ween Sucupira and Ome а, У » U berlandia Uberabinha),
— 1200 (BM, ; Serra de S. José, —
17406 (P. US, W); — Tos ie Macedo . 351 (US); Con-
ceigáo ni Mato Dentro, Rodovia Belho Horizonte,
==
ч
—
=
3otucatú, 18 km N of Botucatu, | 1 km
Sáo Manuel, Cobra 1043 (US). Without state: with-
out locality, 1844, Weddell 2556 (P).
11. Paspalum goyasense Davidse, Morrone &
391, fig. 2. 2001. TYPE:
К de Cabeceiras (ca. 4
Zuloaga, Novon 11:
Brazil. Goiás: 25 km
km E of Goiás—Minas Gerais), cerrado, 1000
19 Nov. 1965, H. S. Irwin, R. Souza & R.
Reis dos Santos 10524 (holotype, MO!; isotype,
US-2529124!). Figure 10.
40—05
internodes 2-3. 6—
Caespitose perennial; culms cm tall,
erect, unbranched, few-noded:
nodes brown, densely pilose. Sheaths 3-10 cm
long, mostly basal, papillose-pilose or hispid, the
margins membranous. Ligules 0.5—0.9 mm long,
membranous, brown, glabrous; pseudoligule absent;
collar pilose. Blades linear, 8-20 cm long, 0.2-0.4
cm wide, mostly basal, flat or with involute margins,
ascendent and rigid, rounded at the base, the apex
acute, densely papillose-pilose on both surfaces,
the margins ciliate. Peduncles long exserted, up to
30 cm long, filiform, pale or tinged with purple.
Inflorescences terminal, exserted, with 2 racemes,
occasionally with a single raceme or with a third
raceme present; main axis up to 3 cm long, flat,
smooth, glabrous, occasionally absent; pulvini gla-
brous, with a squamiform bract; racemes alternate,
ascendent and divergent, ending in a developed
spikelet; rachis of the racemes flat, straight, (2—)3—
7 em long, 1-1.6 mm wide, glabrous, the midnerve
pale to green, the wings hyaline, nerveless, pale to
purple: pedicels short, glabrous: spikelets solitary,
imbricate and arranged in 2 series. Spikelets long-
ellipsoid, 4—5 mm long. 1—1.1 mm wide without the
acute,
hairs, plano-convex to slightly biconvex,
smooth, pilose; upper glume and lower lemma
densely pilose on the lower half with white hairs,
otherwise glabrous, the margins with papillose
hairs, up to 2.5 mm long, reduced toward the apex.
Upper glume as long as the spikelet, acute at the
apex, with or without a short tuft of hairs, membra-
nous, 3-nerved, with one central and two submar-
ginal nerves. Lower lemma glumiform, as long as
the spikelet or slightly shorter, 3—5-nerved. Upper
anthecium long-ellipsoid, 3.2—4 mm long, 0.9 mm
wide, 0.7—1 mm shorter than the spikelet, plano-
convex, membranous to chartaceous; upper lemma
with small papillae regularly distributed, and a tuft
of macrohairs at the apex; lodicules 2, ca. 0.3 mm
long, conduplicate; stamens 3, anthers 2 mm long:
stigma plumose. Caryopsis long obovoid, 2 mm
long, 0.9 mm wide; hilum elliptic; embryo % as
long as the caryopsis.
Distribution and habitat. Endemic to central
Brazil, where it grows in the Federal District and
in the state of Goids, in cerrados between 900 and
1200 m.
Paspalum goyasense is a close species of P. car-
inatum, the latter with filiform leaves, inflorescenc-
es usually with a single raceme, rachis of the ra-
cemes 1.8-2.5(-3.5) mm wide, spikelets with long
—
hairs rising from the lower third, conspicuously sca-
brous on the rest of the surface, and apex of the
upper glume and lower lemma dorsiventrally com-
pressed and flat
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
—“ specimens. BRAZIL. Distrito Feder-
in et al. 10257 (MO, US); 10 km W of
Taguatinga, ruta a Braslandia, Irwin et al. 10654 (MO,
UB, US): Paros da Contagem, ca. 20 km |
Irwin et ч 618 (MO. US): Brasília. Irwin et al. 9658
(MO, US — — Sul, Zuloaga 3835 (SI). —
ae los ality, Glaziou 22441 (G, US, W), — 22443
а US, W): Serra dos Pirineus, 14 lis S de E — de
‚ 1678, 49?W, Irwin et al. 10737 (MO, US): ca. 5 km
N a pe rd. W to Pires do Rio by BR D s el al.
75939 (MO).
N de —
12. Paspalum heterotrichon Trinius, Sp. Gram.
3: pl. 285. 1829-1830. TYPE: Brazil. Without
locality, G. H. von Langsdorff 1829 (holotype.
LE! isotype, K! photo, SI!, US-2855297!). Fig-
ures 15. 16
Paspalum n Schltdl., Linnaea 26: 134. 1854. hom.
illeg. TYPE: Venezuela. Silla de Caracas, —
97 ee
Paspalum — hon Trin. var. dew үте Hack..
Notizbl. . Gard. Berlin-Dahlem 1(8): 328. 1897.
TYPE: ен Monte Furey, 1896, — — (ho-
not seen; isotype, US-2
..
Paspalum ceresioides Carrillo, Revista E. vi 1. Е шу. Nac.
Ante 08. Syn. nov. TYPE: 9 Cuz-
co: Prov. Convención the — 920 m. 14 Abr.
1966. C. Vargas C. 15 (holotype. CUZ not seen:
Sip “U SD.
ntonio 2: 145. 19
Caespitose, short-rhizomatous perennial, cata-
phylls pilose; culms (27-)45-90 em tall, 1-1.9 mm
nodes or un-
=]
diam., branching at the upper
branched; internodes 2-3 cm long on the lower
third and 6-12 cm long on the upper portion of the
culms, terete, hollow, glabrous: nodes glabrous or
pilose, with hairs white, 1-1.5 mm long. Sheaths
(3—)5-12 ст long. glabrous or with the upper mar-
gins pilose. Ligules 0.5 mm long, membranous,
truncate: pseudoligule present. a ring of hairs 3—1
mm long. Blades linear to linear-lanceolate, (4—)5—
15 cm long, 1.5-4 mm wide, the abaxial surface
glabrous. the adaxial one glabrous to sparsely pi-
lose, the margins scabrous, attenuate, the apex
acute. Peduncles (6-) 10-16 em long, terete, pilose.
Terminal inflorescences 6-12 cm long. 2-2.5 cm
wide: main axis (1—)2—4 em long, 0.5 mm diam..
pilose; racemes (1 to 2)3 to 7, (27)3.5—6.5 cm long.
alternate, ending in a naked point; pulvini pilose:
rachis of the racemes 3.54 mm wide. acuminate.
glabrous, green at the middle portion and ferrugi-
e
nous al the wings, wings nerveless; pedicels 0.4
mm long. hispid, laterally inserted in the spikelets:
spikelets solitary, arranged in 2 series. Spikelets
ellipsoid, 2.3-3 mm long, 1-1.2 mm wide. pilose.
upper glume and lower lemma subequal, with an
annular thickening at the base. Upper glume as
long as the spikelet, hyaline, acuminate, 3-nerved,
with basal hairs Y the length of the spikelet and
papillose-pilose hairs at the margins, up to 3 mm
long. Lower lemma membranous, acuminate, 3-
nerved, with long hairs at the upper margins, oth-
erwise glabrous. Upper anthecium ellipsoid, 1.5—
1.8 mm long, 0.5 mm wide, glabrous. papyraceous:
upper lemma with a tuft of macrohairs at the apex,
and bicellular microhairs and silica bodies all over
the surface; upper palea with prickles and silica
bodies on the upper portion; lodicules 2, 0.1 mm
long. rounded at the apex; stamens 3, anthers 1.2
mm long: stigma plumose. Caryopsis not seen.
Iconography. Chase, Contr. U.S. Natl. Herb.
28(1): 18, fig. 2. 1929. Swallen, Ann. Missouri Bot.
Gard. 30: 93, fig. 18. 1943. Rodriguez, Ernstia 8(2—
3): 38. 1998.
Distribution and habitat. Paspalum heterotri-
chon grows in Mexico, Central America (Honduras
and Panama), and the Caribbean (Haiti and the Do-
minican Republic); in South America it is found in
Colombia, Venezuela, Peru, and central Brazil, usu-
ally in open fields on limestone or rocky soils, be-
tween 300 and 2550 m elevation.
Paspalum heterotrichon shares with P. petrense.
P. trachycoleon, and P. phyllorhachis ellipsoid
spikelets, thickened at the base. upper glume hy-
aline, and rachis of the racemes winged. Paspalum
petrense differs by the long-acuminate spikelets and
lower lemma with a similar pilosity pattern of the
upper glume, i.e.. with long hairs at the base of the
surface and the margins. Paspalum trachycoleon is
distinguished by the paired spikelets, while P. phyl-
lorhachis has glabrous spikelets.
The specimen Macedo 1121, probably a poorly
developed plant, presents smaller spikelets than
typical, 1.7-2 mm long, and culms 25-30 cm tall.
Internodes are occasionally markedly branching,
З em long and thinner, up to 0.6 mm in diameter.
In P and US (2855297) there is a specimen, Rie-
del s.n., that probably belongs to the type collection
1978,
ments on Langsdorff and Riedel collections).
of P. heterotrichon (see Renvoize. for com-
BRAZIL. Distrito Estar:
al: Reserva Ecológica do IBGE. Oliveira 719 (SI).
without locality, Glaziou fe (G, К,Р, W).
Mir las Gerais: Mun, ltu-
Representative —
Mato Gros-
so: Cuiabá,
bitaba.
cality, Riedel 966 (G, 51). s.n.
к Ol — A. Magdalena: Santa Marta,
(P. US-2 855297).
Smith 143 (BM,
LOL, G, МО, P). DOMINICAN REPUBLIC. Azua: San
т тө La Vieja, Cordillera Central, Ekman 13404
(US). Santo Domingo: Cordillera Central. Prov. de la
Vega, Costanza, Bajo de La Angostura. Ekman 14018
372 Annals of the
Missouri Botanical Garden
==
m
2
] W 7
WS
\
ш
SX
RRS
ү au
я] ^ М
Figure 15. Paspalum i ва Habit. —B. Detail of ligule. —C. Portion of the rachis, —D. Spikelet,
dorsal view. —E. Spikelet, ventral view. Е Upper anthecium, dorsal view. —G. Upper — ventral view. —
H. Upper palea ih lodic m з. —l. Ci aryopsis, scutellar view. —J. байбы, hilum view. (A, В, D-J based on Malme
1562b, US; € based on Tovar 4176, US.)
p
—
ocotzi X-62.3 (US); Mun.
m, 24 Mayo 1989, Dias et al. 3493 (MO).
idse 3082 (К. MO); prope Colonia Tovar, Fendler 1698 (G,
taume V0530 (MO
Volume 89, Number 3 Denham et al.
2002
373
Paspalum subg. Ceresia
|
|
|
|
|
|
|
|
|
A P. heterotrichon |
e 7 lanciflorum |
|
|
|
\ |
с \ \ \
\ \ |
\ \ \
\ \
\ \
\ \ |
\ \ \ |
\ y \
\ p pi |
\ \
non \ \
Q0 200 400 600 son Bs km \ \ |
— E не ý
u 100 200 — A00 St ТҮ miles \ \
Prepared by Hem K Rypkema J |
— — — e =
Figure 16. Distribution of Paspalum heterotrichon and P. lanciflorum
МО. US). HAITI. du Nord: vicinity of St. Michel Че 13, Paspalum humboldtianum Flüggé, Gram
í
; - V er, Paspe : . . as "Paspalus
Ennery, . Leonard 8953 (MO), 8957 (US): Frouin, eruptive Monogr | "id P 61 1810, da "1 sis ni
hills above C hapelle Jubilée, Ekman 2401 (G, US); Massif humboldtianus." Panicum humboldtianum
des Matheux. 26 Oct. 1924, Ekman 2274 (US). HON-
(Fliiggé) Kuntze, Revis. Gen. Pl. 3(2): 361.
DURAS. Comayagua: 4 km from San Isidro on the road 1898. TYPE: Ecuador: Puembo, А W. H. A.
to Pane, S side of rio Tula Gorge, Moran 5505 (MO). El von Humboldt & A. J. A. Bonpland 3104 (ho-
Paraíso: near km 2, road to Yuscarán, 5 Nov. 1951. Swal- | . B-W! pl SI!: is BMI. PLUS
len 11354 (MEXU, MO, US). Francisco Morazán: re- огуре, : photo, 515; isotypes, Mq
gion of Las Mesas, Swallen 10810 (MO, US). 1 1401 (US).
601340! photo, SI!). Figures 14, 17
XICO. Chiapas: El Pozo to Oxchuc, Hernández Xol-
Villa Corzo, above Colonia Vi-
Guerrero on road t Finca Cuxtepec, Breed
=
Paspalum distic к Kunih, in H.B.K., Nov. Gen.
P
te love & Sp. 1: 86. 1816 TY ;olombia: Mesa de WI
pe 54596 (MEXU, US). PANAMA. Chiriquí: near et Ibagué, Oct 1. A. von Humboldi & A. J. /
El Boquete, Hitchcock 8297 (SI, US). PERU. — Bonpland s.n. ‚ Mee P!; isotypes, B!, —
Luya Province, C spa и Cedro, 2450-2550 2855210!)
Huanc s em Paspalum ciliatum Kunth, in H.B.K., Nov. Gen. Sp.
Huancavelica, Distrito de v * 'olpa, entre Marcavalle pl. 24. 1816, hom. Шер. Paspalum Bepharophorum
y Quintabamba, Tovar 4176 (US). VENEZUELA. Aragua: Roem. & Schult., Syst. уы 2: 292. 1817
km S of Alto de Choroní along road to Maracay, Dav- Colombia: Ibagué у, Valle e Caravajal, Monte Quin-
diu, Oct. Н. A. von Humboldi & A. J. A. Bonp-
d s.n. holotype PL isotypes, B!, ся 2855275!).
Panicum obtectum J. Presl, — Hae nk. 1: 301. 1830.
Tricholaena obtecta (J. Presl) Ё — ex Hemsl.,
Lara: Distr. Palavecino, Quebrada La Mata, towards
side of the Parque Nac. Terepaima, Burandt & Gan-
374 Annals of the
Missouri Botanical Garden
=
— —
|
ES
Figure 17. Paspalum лшн gi n —A. Habit. —B. Detail of ligule. —C. Portion of the rachis. —D. Spikelet,
dorsal view. —E. Spikelet, ventral view. —F Upper anthecium, — view. —G. Upper anthecium, ventral view. —
H. Upper palea, lodicules, and gynoecium. (Based on Zuloaga et al. 5887, SI.)
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
375
iol. Centr.-Amer., Bot. 493. 1885, comb. illeg.
fluens obtecta (J. — Fourn., Mexic. Pl.
1886. TYPE: Mexico. Without locality, Z P.
X. Hank x s.n. 1. (holotype, PR not seen; isotypes, US-
2942530!.
Paspalum ul Hack.. in Stuck., Anales Mus. Nac.
juenos / 11: 63. 1904. Paspalum ps
um Flüggé var. stuckertii (Hack.) Hack.. Allg. Bot.
Z. Syst. 12: 97. 1906. TYPE: Argentina. Córdoba:
Dpto. Punilla, Sierra Chica de Córdoba, en la punta
de la — del Salto, del Ayo. Ochoa, 1000 1 m,
l T. J. V. Stuckert 13023 (holotype,
ires
-BA AL CORD!, G!, MO, US-80012! —
SI! US-557978)).
Paspalum — rum yc Contr. U.S. Natl. Herb. 24:
443. 1927. Syn. TYPE: nd Loja: — n
Loja ме Sim — 2100 m, 6 Sep. 1923, 4.
d 21495 (holotype, Us. 1164798! photo. B
vpe. 5 not s
fadus humboldrianum ir var. elegantissima Bee-
Пе. Phytologia 52: 15. 1982. — nov. TYPE: Mex-
ico. коне ) km — Los wr os, San José de
Andrade, 29 Sep. 1980, A. Бойе M-5962
AS
(holotype, not located).
-
trj
Short rhizomatous perennial, with leafy rhizomes:
culms (30-)50-100 em tall, 0.2-0.3 cm diam..
erect or decumbent, lignified at the base, branching
at the lower and middle nodes or unbranched: lower
internodes 0.51 cm long, the upper ones 4-15 cm
long. terete, hollow, glabrous or shortly pilose at the
distal portion; nodes glabrous to pilose. Sheaths 4—
l cm long, longer than the internodes, overlap-
ping. densely to sparsely papillose-pilose. with ca-
ducous tuberculate hairs at the upper portion, the
margins ciliate. Ligules 1.2-2.5 mm long, membra-
nous, brown, glabrous; pseudoligule absent: collar
5-2] cm
glabrous
Blades lanceolate,
flat,
to sparsely pilose or hirsute, with long, tuberculate.
glabrous or pilose.
long. 0.5-1.8 cm wide, herbaceous,
caducous hairs, green glaucous, shortly pseu-
dopetiolate, the apex subulate. Peduncles exserted,
up to 30 em long, terete, glabrous. Terminal inflo-
rescences exserted, 8—12 cm long. 3-5 cm wide:
main axis 2-7 cm long, flat, with longitudinal
nerves on the wings, scaberulous; racemes 2 to 6
ascendent, alternate and distant, slightly divergent
from the main axis, ending in a naked point, oc-
casionally in a spikelet; pulvini with a long tuft of
white, rigid hairs; rachis of the racemes (3—)5-9 cm
long, 1.2-1.5 mm wide, flat, shortly winged, green
to purple. glabrous, scaberulous on the margins:
pedicels unequal, up to 1.2 mm long, scaberulous:
spikelets paired, and imbricate in 4 series, occa-
sionally the lower spikelet aborted. Spikelets long-
ellipsoid, 3—4 mm long, 1.2-1.4 mm wide, acute.
plano-convex, pale or tinged with purple, the mar-
gins with long hairs: upper glume and lower lemma
as long as the spikelet. Upper glume 3-nerved, hy-
aline, with plicate, corky margins covered with rig-
2-3.2 mm long.
the rest of the surface sparsely pilose. Lower lemma
id. radiate and tuberculate hairs.
3-nerved, membranous, glabrous, scaberulous at
the upper portion, depressed at the base. Upper
anthecium ellipsoid, 2.6-3.2 mm long, 1-1.2 mm
wide, membranous, pale, shiny, 0.4—0.6 mm shorter
than the upper glume and the lower lemma: upper
lemma 5-nerved, finely papillose with silica bodies.
bicellular microhairs and prickles at the apex: up-
per palea papillose, with bicellular microhairs at
the distal portion; lodicules 2, 0.6 mm long, trun-
cate, stamens 3, anthers 1.6
conduplicate: mm
long. Caryopsis not seen.
Common name. “Canutillo” (Venezuela).
Chromosome number. n = 10 (Hunziker et al..
1998); п = 20 (Saura, 1948); 2л 40 (Tateoka.
1902).
Distribution and habitat. Found from Mexico
and Central America to Bolivia and central and
northwestern Argentina. |t. grows on mountain
slopes. on loose. rocky to sandy soils: it is also
common at margins of highways and in forest edges
or modified areas. between 400 and 2900 m ele-
valion.
Paspalum humboldtianum is related to P. poly-
phyllum and P. buchtienii by the upper glume with
corky margins and covered with radiate hairs, the
rachis usually ending in a spikelet. Paspalum po-
lyphyllum differs by the narrow (0.1-0.6 em wide).
to linear-lanceolate
not pseudopetiolate, linear
blades. rachis pilose and with anastomosed veins.
spikelets with the upper glume with unequal mar-
ginal hairs. and lower lemma pilose on the distal
margins and the upper portion of the surface. Dif-
ferences with P. buc htienii v were summarized in the
treatment of the latter spe
Paspalum — жанан has spikelets
dark
brown: however, pale spikelets and anthers were
tinged with purple and anthers purple or
observed in several specimens, such as Zuloaga et
al. 5885.
The type material of P. humboldtianum var. ele-
gantissima was nol located; it should be mentioned
that no herbarium was cited by Beetle (1980) when
However,
describing the species. the diagnosis al-
lowed us to conclude that this variety is a synonym
of P. humboldtianum.
hen describing P. soboliferum based on a sin-
1927) distin-
guished the species from P. humboldtianum mainly
gle collection from Ecuador, Chase (
by the shorter, more slender, and ascending ra-
cemes, spikelets 2.8-2.9 mm long, with short hairs
toward the margins. Therefore, we considered the
376
Annals of the
Missouri Botanical Garden
type collection a poorly developed plant and placed
P. soboliferum as a synonym of P. humboldtianum.
Spikelets with proliferous bracts were observe
in several specimens from Mexico, such as Müller
2036, Arséne 2813, and Schaffner 136.
Representative — ARGENTINA, ооа
чү Andalgalá, Las Payvas, bu 1765 (LIL,
US). Córdoba: Ascochinga, Burkart 10211 (SD); rae
Send Sierra Chica, Stuckert 14159 (CORD, G, MO, SI).
Jujuy: Río Perico, Burkart & Troncoso 11014 (SI); Dpto.
Volcán, — : = 5885 (SI). La Rioja:
а: Dpto. € al San Luis,
ees Cuesta de
. ‚ US). San Luis: bod
sdernera, San José del Morro, Ea. La Morena, Boel-
ске & Moore 16646 (SI). Tucumán: Taff del Valle, Bur-
kart 5335 (SI); Dpto. Burruyacú, camino a Cerro Medina,
Villa Carenzo et al. 1818 (LIL). BOLIVIA. Chuquisaca:
Sucre by the exit to Tarabuco, Wood 7852 (US). Cocha-
bamba: Morochata, Cárdenas 3410 (LIL); W of Cocha-
bamba, — rie ) (US). La Paz: viciniis Sorata,
Mandon 125: M, C, P. US, W). Santa Cruz: 10 km E
of Samaipa, ^ ^nvoize « Cope 4038 (MO); Buena Vista,
Steinbach 6618 (BM, G, US). Suere: Sucre, Cárdenas 515
US). Tarija: Camargo, Fiebrig 3082 (G, LIL, SI, W). CO-
LOMBIA. Antioquia: San Jerónimo, Aralar 629 (MA,
US). Boyacá: without locality, Saravia 4006 (COL). Cau-
ca: Chisquío, Finca Los Derrumbos, Asplund 10498 (G,
US). Cundinamarca: La Esperanza, Cuatrecasas 3001
(MA). Huila: 6 km SE of Altamira along road to Florencia,
Davidse et al. 5597 (COL, MO). Norte de Santander:
vicinity of Pamplona, Killip & Smith 19770 (US). Prov-
idencia: Pico-casabaja, /drobo 11636 (COL). Santander:
Zapatoce: Fassett 25502 (COL, US). Without departa-
ment: without locality, Mutis 5537 (US). — Valle
de San Juan, Cerro del Real Mina del Sapo, Echeverry
1251 (COL). Valle: near Sevilla, Barclay ‘ Juajibioy
5703 (MO, US) Cordillera Occidental, E ope, : > 1,
Loma Los Cristales, Cuatrecasas et al. 25702 (€ OL. US).
COSTA RICA. Cartago: zu Juan Norte, Pohl et uil
11434 (US). Guanacaste: 2 km E of Liberia, Pohl et al.
11327 (US). ECUADOR. Azuay: vicinity of Cue nea,
along Río Milchichi, about 5 km N of Cuenca, Camp E-
2743 (G, US). Chimborazo: Chunchi-Alausi, Acosta Solís
21454 (US). Loja: halfway between Loja and Catamayo,
Grignon 84257 (MO). Oro: between Portovelo and E
Tambo, Hichcock " 306 (US). Pic hincha: Tumbaco, As-
plund 6531 (G, P, US). Tunguragua: Ambato, ue
21710 (US); UR Asplund 7593 (G, P, US). 7
VADOR. Ahuachapán: P.N. El D Sermeño s.n.
(MO). San Salvador: San Salvador, Calderón 1152 (US
GUATEMALA. Alta Verapaz: 5 mi. S of Cobán is
Hwy. CA 14, Thomas s.n. (MO). Santa Rosa: Cerro Re-
dondo, Heyde & Luz 6271 (6). Sololá: Volcán San Pedro,
Steyermark 477177 (US). HONDURAS. Francisco Mor-
azán: vicinity of El Zamorano, Swallen 10776 (US).
Olancho: Mun. La Unión, 5 mi. E of La d along
road to Olanchito, Davidse et al. 35448 (MO). M
Chiapas: 22 km NE of Motozintla on hwy. to с nango.
Stevens & Martínez 25701 (SI). Colima: Alzada, Hitchcock
7055 (US). Guanajuato: ca. 8 km W of San Felipe on
Cerro del Fraile, Sohns 427 (U: : Montes de
Oca, Vallecitos, Hinton 11400 (US). Jalisco:
jomulco de Zuñiga, Cerro Viejo region, 3:
SSW of Guadalajara, Machuca Nuñez 4640 (MO). Mi-
Tumbaya:
choacán: Mun. Apatzingan, Leavenworth & Leavenworth
1649 (US). Morelos: Cuernavaca, Hitchcock 6837 (US).
arit: La Yesca, 23 km al SE
Tenorio & Flores 16710 (MO). Oaxaca:
Mun. Santa María Ecatepec i :
Santa María Zapotitlán, Flores Flanco et al. 3643 (MO).
W of San Luis Potosí, Reeder et
3 (US). Sonora: Ta popa, Sierra Saguaribo, —
etal. 19307 (US). Tamaulipas: Mun. Villa de Casas, cen-
tral e of the Sierra de Tamaulipas, Martínez & ee
F-1953B (US). Veracruz: Zacuapán, — 2002 (US).
Zacatecas: Mun. Valparaiso, 1 km a San Francisco,
Rzedowski 176. 36 (US). NICARAGUA. Nueva Segovia:
Santa María de los Pinos, 5 km NE de Dipilto, Stevens &
Montiel 24761 (MO). PANAMA. Chiriquí:
Partch 69-91 (MO): roadside near El Hato
Stockwell & Aiello 893 (MO, SI). PERU. Gasca Prov.
Chachapoyas, 5 km W of C bac ‘hapoyas on road to Caclic,
Hutchison & — — (G, MO, P, US
Challhuanaca, Ayma . Núñez 7175 (M
Chumbes, Balls 6023 (BM. US). Cajamarca: Prol Chota,
Yamaluc, a 11 km sobre la carretera Cochabamba—Huam-
bos, Sánchez Vega 2341 (SI). Cuzeo: Prov. Convención,
Amaybamba, Marin 851 (US). Huancavelica: entre Iz-
cuchaca y Acoria, Tovar 1005 (US). Lambayeque: Prov.
Chiclayo, Cerro Reque, — Quiroz 2322 (MO). Lima:
entre ap dope y Cañete, Ferreyra 5373 (US). VENE
ZUELA. Aragua: Distr. Ric 'aurte, ca.
Colonia. Tovar along the Colonia Tovar-La Victoria Hwy.,
Davidse & Miller 28159 (MO). Distrito Federal: 5.5 km
down poe to Carayac a, Davidse & lI ^tt 4073 (MO).
Lara: Jim S of 9°42
09°38 ү pons үз — 21370 (M a Mérida: Se-
4 (US). Miranda: Los Teques, Bur-
(S1). Táchira: Mun. Andrés Bello, Loma
Aldea Salomón, Tamayo 2281 (US).
>
—
=
>
-
7 5
«d
=
—
3
=
=
2
Blanca,
14. Paspalum imbricatum Filgueiras, Bradea 3:
153, fig. 2. 1981. TYPE: Brazil. Mato Grosso:
Mun. Sidraléndin. Rod. BR-163, orla de brejo,
27 Oct. 1970, G. Hatschbach 25281 (holotype,
US-2705820! photo, SI!). Figure 25.
0.3—
un-
Caespitose perennial; culms 1-1.5 m tall,
0.5 em diam., herbaceous 1
» rigid, erect,
branched: internodes 15-30 cm long, slightly com-
pressed, hollow, glabrous, pale or tinged with
purple; nodes brown, glabrous. Sheaths 15-25 em
long. densely hirsute, with papillose-pilose hairs at
the distal portion, glabrous at the base, the margins
membranous, shiny on both surfaces. Ligules mem-
1-1.5(-2) mm long,
pseudoligule a ring of hairs up to 1.3 mm long.
35-60 cm long, 0.6-1 cm
branous, brown, glabrous:
Blades linear-lanceolate,
wide, acute, pungent, mostly basal, flat, rigid, as-
cendent, the basal ones long pseudopetiolate,
densely hirsute, with tuberculate hairs, on both sur-
faces. Peduncles up to 40 em long, terete, glabrous,
pale or purple. Terminal inflorescences exserted,
10-15 cm long, 8 em wide; main axis 1.5-5.5 cm
long, flat; racemes (2)5 to 7(1 1), (5-)10-14(-16) em
long, ascendent, slightly divergent from the main
Volume 89, Number 3
2002
Denham et al.
rs Ceresia
377
axis, alternate, ending in a naked point: pulvini
shortly pilose; rachis of the racemes 2.4—2.8(—3)
mm wide, flat, the margins membranous to hyaline,
nerveless, glabrous, green, shortly winged. wings
purple; pedicels short; spikelets solitary, densely
imbricate in 2 series. Spikelets broadly ovoid,
5.5 mm long, 2.8-3 mm wide, plano-convex, gla-
brous, pale, upper glume and lower lemma sube-
qual, or the lower lemma slightly longer, cordate,
the apex acute to almost rounded, the margins
winged. Upper glume papyraceous, corrugate in the
lower half, 5-nerved with 3 central nerves reaching
the apex. Lower lemma glumiform, 5-nerved, cor-
rugate in the lower half. Upper anthecium long-
ellipsoid, 3.8-4 mm long, 1.2 mm wide, plano-con-
vex, membranous, pale, pilose toward the apex.
finely papillose; upper lemma with small papillae
evenly distributed all over the surface. with macro-
hairs and bicellular microhairs at the apex, hispi-
dulous toward the margins and in the lower portion:
upper palea hispidulous toward the apex: lodicules
2. ca. 0.4 mm long, conduplicate, hyaline: stamens
3. anthers 2-2.5 mm long; stigmas 2. plumose. lat-
eral. Caryopsis nol seen.
Brazil,
where it is found in cerrados in the Distrito Federal
Distribution and habitat. Endemic to
and the states of Mato Grosso and Goiás, between
950 and 1250 m elevation.
Paspalum imbricatum is a member of section
Pectinata, sharing cordate upper glume and lower
easily distin-
lemma, the margins winged; it is
guished by the glabrous spikelets.
15. Paspalum lanciflorum Trinius, Sp. Gram. 3:
pl. 286. 1829-1830. TYPE: Brazil. M
alo Gros-
Cuiabá, 1829, H. von Langsdorff s.n
(holotype, LE!; isotype, US-80075!). Figure
16.
uc contractum Pilg., E^ — st. 29: 700.
898. TYPE: Colombia. M Janos de San
жй — ү 68, M. A. pu po A ue В!;
isotype, US-2942134!
Р ке ес chino Mez, Bot. Jahrb. Syst. 50, Beibl
. 1921. TYPE: Brazil. fis Bra
iy ern
— мш 1909, E. H. Ле —
designated by Toda WICZ (1990: йч BL isolecto-
types, K!, US-2854675! photo, US).
ын — Swallen, Fieldiana, Bot. 28: 24. fig.
19 TYPE: ека Bolivar: Gran Sabana,
e Kun г ara—Paru, in d of Rio
Kuke пап, S of Mount Ro aima, 1065— ym
A. Steyer mark 59090 — US-
19116601: Isotype.
Р pun aureolatum Sells n, Fie Шала, Bot. 28: 22, fig
1951. TYPE: l'erritorio Federal An
zonas: summit of Cerro Duida, 1025—1200 m, 2 Ben.
1 n
Venezuela.
1944, Steyermark
J 58234 (holotype, US-
1911652! photo, SI!
—
Caespitose, shortly rhizomatous perennial; culms
45-120 cm tall,
14 cm long, glabrous, brown: nodes glabrous to pi-
erect, unbranched; internodes 9—
lose. brown, compressed. Sheaths 9-17 cm long,
densely hirsute on the distal portion, otherwise
sparsely pilose or glabrescent, the adaxial surface
1.5-3.2 mm
long, glabrous, brown; pseudoligule a ring of hairs
shiny, brown. Ligules membranous,
up to 0.8 cm long. Blades linear-lanceolate, 7-35
em long, (0.3—50.5-1(-1.5) em wide, flat or with
involute margins, densely papillose-pilose, with tu-
berculate hairs up to 7 mm long, the midnerve con-
spicuous on the adaxial surface. Peduncles long-
exserted, up to 40 cm long, terete, glabrous.
Terminal inflorescences subdigitate, yellow, occa-
sionally with a papyraceous bract, up to 4 mm long,
at the base; main axis absent or up to 1.5 em long,
33 to O(11),
ascendent, 5—15 em long, subconjugate and ending
—
glabrous, flat; pulvini pilose; racemes (
in a naked point; rachis of the racemes (4—)5—7 mm
wide, winged, green at the center, the midnerve
white, with purple, membranous and glabrous mar-
gins: pedicels short, glabrous. flat: spikelets soli-
ary, imbricate in 2 series. Spikelets long-ovoid,
—
(4.8—)5.6—7.2 mm long, 1.2-2.5 mm wide, plano-
convex, acuminate, pale, pilose. Upper glume as
long as the spikelet, membranous, rounded and
densely pilose at the base, the basal margins pilose,
with hairs up to 2 mm long, otherwise glabrous, 3-
nerved. Lower lemma slightly shorter than the up-
per glume, lanceolate, 3-nerved, densely pilose at
the basal margins, the median portion with con-
spicuous, tuberculate hairs 2.5-3 mm long. Upper
3.24.7 mm long, 0.9-2
wide, membranous, pale, stipitate, the stipe 0.3 mm
anthecium ovoid, mm
long: upper lemma finely papillose, with bicellular
microhairs and silica bodies at the apex, otherwise
lodicules 2,
cale; stamens 3, anthers 1.6 mm long; stigma lat-
elabrous: ca. 0.4 mm long, condupli-
eral. Caryopsis not seen.
Iconography. Chase, Contr. U.S. Natl. Herb.
28(1): 26. fig. 7. 1929 (under P. contractum). Jud-
ziewicz, Fl. — 469. fig. 83, E. 1990
Common name. “Pama” (Venezuela).
Distribution and habitat. Panama, Colombia,
Guyana, Surinam, Venezuela, Brazil (Roraima,
"J
зага, Maranhão, Mato Grosso, Minas Gerais, Goiás,
and Distrito Federal), and Bolivia; it grows in sa-
vannas subjected to fire, on limestone, sandy, or
rocky soils, between sea level and 1200 m
Paspalum lanciflorum belongs to section Pectin-
ata; within this section it is related to P. cachim-
378
Annals of the
Missouri Botanical Garden
with
branched, geniculate culms, rachis of the racemes
boense, which differs by the annual habit.
2-3 mm wide, spikelets 4—4.3 mm long, and upper
anthecium 2.2-2.5 mm long. Also, P. lanciflorum
is distinguished from P. pectinatum and P. cordatum
by the glume densely pilose on the lower portion,
not cordate, and upper anthecium stipitate, 4—%
the length of the spikelet.
This species is quite variable in its overall mor-
phology. The specimens Wurdack & Guppy 27, Tate
157, and Hahn et al. 5592b are approximately 40—
50 em tall, with all blades basal. On the other hand,
Kuhlmann 1674, Glaziou 20084, 22544, and Zu-
loaga et al. 4413 are more robust specimens, ca.
120 cm tall, with blades evenly distributed along
the culms. It is noteworthy to mention the presence
of densely papillose-pilose specimens, e.g.. Zuloa-
ga et al. 4413, Tamayo s.n., Egler & Raimundo
1262, and others, as sparsely pilose, such as Wil-
liams 13088 and Maguire et al. 32038.
Three to six racemes per inflorescence are fre-
quent in this species; nevertheless, one to thr
racemes are found in Swallen 4029, Williams
13088, Wurdack & Guppy 27, Gifford G-101, Oliv-
eira & da Silva 700, Tamayo 3198, Ramirez 961,
and Heringer et al. 4615; the specimen Echeverry
et al. 2:
Spikelets are small, 4.8—5 mm long. 1.7-2 mm
wide in Davidse et al. 22647 and Blydenstein 1468.
Chase (ined.) distinguished P contractum. from
P. lanciflorum by the number of racemes (1 to 3 vs.
3 to 6) and the relative length of the lower lemma
(as long as the spikelet vs. shorter than the spike-
let). Our analysis of herbarium material did not al-
is 11 racemes per inflorescence.
low us to distinguish between both taxa.
According to Judziewicz (1990), the specimen
Pohl 1454 (seen in B and W) is a syntype of the
species; nevertheless, we examined at LE a speci-
men collected by von Langsdorff (see type materi-
al), and there are no elements that suggest Trinius
had cited two syntypes when describing the spe-
cies.
VIA. Santa Cruz:
Noel
Representative specimens. BOLI
Vel
— elasco, Parque Nacional Kempff Mercado,
Campane nto Las Gamas, Killeen et al. 4929 (MO). BRA-
ZIL. Distrito Federal: с Baci ча do Rio S. Barto-
lomeu, He — et al. 46 erra de Arruda, Glaziou
4 (K, P, US); Serra Dorada Pohl 1454 (G). Goiás:
Mur n. Couto Megülhaes, 16 km W of Piqizeiro along GO-
70, Plowman et al. 19132 (US). ode — Mor-
ro antes do Rio Tem, Oliveira & da Silva 700 (S1): Caro-
lina to San о ve Balsas. Swallen 402 ds (US). Mato
Grosso: ca. 270
de Xavantina, Gifford G-101 (US):
entre Bardo de ee ma e Utiarity. Kuhlmann 1674
(US). Minas Gerais: Ribeiráo Taquaruc u, Glaziou
( ‚Р, US). Para: Tapajós, Rio Cururú, Egler & R
mi 1262 ( US). Roraima: Aldo ча Tucham Paulo, para
a base do Roraima, Rondon 5. A ea 2454780); Mt. Ro-
raima, Paulo, Tate 157 (US OMBIA. Cravo Norte
Ara uca, al S del Río Meta, a 1468 (US). Meta:
O de La Macarena, vía a C Е — & Ma-
ld 6886 (Sl): carretera de Bella Vista a Piñalito,
Echeverry & Taranto 2312 Mun.
Puerto Carreño, ' Amat & — 3 (COL).
GUYANA. Potaro- Siparun ni Ra egion, ahn et al.
oe (US); Annai Pus — 923 (US). PANA-
A. a: ca. 7 mi. of Minas, D'Arcy &
Antonio 13527 (MO). SURINAM. Sipaliwini savana area
on Brazilian анвар Oldenburger et al. 5 (US). VE
zonas: Canaima, R. Carrao, García Barriga
(COL): Santa Barbara savanna, m of Ríos
Ventuari and. Orinoco, Maguire А al. 8 (US); alre-
dedor del bajo Río Cataniapo, P жее ис cho Williams
13088 (US). Anzoategui: Dist. Bolívar, S of El Zamuro,
Fila El Purgatorio, Davidse & — 19314 (MO). Bo-
livar: El Pauji, Liesner 1935 0 (SI); pié de la Roca to
Guayaraca, Davidse & Huber 22647 (МО); Gran Sabana,
Vía Kavanayen, El ч — 961 (MO): Hato Santa
Teresa, Tamayo — Hato La Vergareña, Wurdack
& Guppy 27 (US 1 al S de San Ignacio de Yuruaní,
Gran Sabana, Zuloaga et al. 4413 (K, SI, US).
Herrer
Z`
la
16. Paspalum longiaristatum — & Fil-
gueiras, Novon 3: 129, fig. TYPE:
Brazil. Goiás: Mun. Ode ‘Me edo,
14°18'5, 48?23'W, 13 Арг. 1992. T. S. Fil-
gueiras 2277 (holotype, IBGE!; — B not
seen, BM!, FLAS not seen, ICN!, ISC not seen.
‚ MEXU-62.8764!, MO!, NY not seen, Р not
SI!, RB not
UB not seen, US not
Figure 23.
seen, SP not seen, R not seen,
seen, UFG not seen,
seen).
Caespitose annual; floriferous culms 15-36 cm
tall, 0.8—1 mm diam., branching at the lower nodes:
internodes terete, hollow, glabrous; nodes dark, pi-
lose. Sheaths 3-7 cm long. papillose-pilose, the
margins smooth. Ligules membranous-ciliate, the
membranous portion 0.5-0.8 mm long, cilia 0.2—
3.5-8.2 cm long, 1-2
mm wide, attenuate at the base, papillose-pilose on
0.8 mm long. Blades linear,
both surfaces, more densely so toward the base.
Peduncles terete, 0.5 mm diam., glabrous or with
papillose-pilose hairs. Terminal inflorescences ex-
serted, 5-8 cm long: main axis 1-2.5 cm long, pa-
pillose-pilose, ending in a naked point; racemes |
to 2(4), 2-7 cm long, alternate, arcuate, ascendent
and divergent; pulvini pilose; rachis of the racemes
membranous, 4—6 mm wide, winged, hyaline, and
ciliate at the margins, the adaxial surface with pa-
pillose-pilose hairs, the abaxial surface glabrous;
pedicels 0.1-0.2 mm long, hispid; spikelets soli-
tary, imbricate in 2 series. Spikelets ellipsoid, 1.8—
2.2 mm long. 0.5-0.6 mm wide, pilose, brown or
tinged with purple, shiny. Upper glume hyaline, 3-
nerved with one central and two marginal nerves,
pilose at the base, the hairs reaching % the length
Volume 89, Number 3
2002
Denham et al.
Paspalum sib Ceresia
of the spikelet. the margins pilose with ascendent
hairs, awned, awn 6-12.2 mm long, scabrous. Low-
er lemma glumiform, slightly shorter than the upper
glume, membranous, 3-nerved, awned, awn 0.3-2
mm long, scabrous. Upper anthecium ellipsoid,
1.5-1.8 mm long, 0.7-0.8 mm wide, membranous.
shiny: upper lemma 5-nerved, with prickles, bicel-
lular microhairs and silica bodies at the upper por-
tion; upper palea 2-nerved: lodicules absent: sta-
mens 3, anthers 1.4-1.8 mm long. purple al
maturity: stigma plumose. Caryopsis 1-1.2 mm
long, 0.1—0.5 mm wide; hilum punctiform. basal,
embryo more than half the length of the caryopsis.
Distribution and habitat. Endemic to central
Brazil.
cerrados. in serpentine, rocky soils at 1000 m.
in the state of Goiás, where it is found
Paspalum longiaristatum differs from P. biaris-
tatum by the perennial habit, spikelets 3.8—1.5 mm
long. awn of the upper glume 4—7.1 mm long. and
awn of the lower lemma 3.84.5 mm long of the
latter taxon.
Representative specime ns. BRA Goias:
а. 15 km N of Niquelándia, E el v 144 (
gee а 18 km de Niquelándia. тая б т
SI. y Filgueiras 2780 (MO, . Oliveira et al. 638
M A 23 ao 3 da estrada e 3 rra a direita da mina de
niquel. Fonseca et al. 227 (MO, SD): Mun. Alto Paraiso.
14°10'S, 98°47 i Filgueiras & Oliveira
— ‘do.
О); P»
^
FS
Vale da Lua.
3292 (US)
17. Paspalum malmeanum Ekman. Ark. Bot.
10: 12. tab. 4. fig. 3. 1911. TYPE: Brazil. Mato
Santa Ana da Chapada, 31 May 1903.
A. Malme s.n. (holotype, S not seen: iso-
-80089!). Figure 23.
gw
* Us
Caespitose, shortly rhizomatous perennial: culms
60-90 cm tall.
long. purple. terete, hollow, striate; nodes glabrous.
few-noded: internodes 6—10 cm
Sheaths usually longer than the internodes, slightly
compressed, glabrous, pale or tinged with purple,
shiny. the margins ciliate or glabrous. Ligules mem-
branous, 0.2 mm long. brown, glabrous: pseudoli-
gule a ring of hairs up to 1.2 mm long. Blades
linear, 10—45 cm long, 0.1 em wide. involute. pa-
pillose on the adaxial surface. glabrous and smooth
on the abaxial surface, the margins papillose-pilose
toward the base with hairs up to 6 mm long. oth-
erwise glabrous, the apex subulate. Peduncles ex-
serted, terete. up to 35 em long. glabrous. Terminal
inflorescences exserted; main axis absent: pulvini
with a tuft of short hairs and occasionally with 2
small bracts on the basal pulvini; racemes 2, con-
jugate. 7—14 em long. occasionally with a single
raceme present; rachis of the racemes 1.2-2 mm
wide, ending in a naked point, purple or with the
margins red-tinged, glabrous, the wings flat: pedi-
cels subterete, glabrous to short pilose at the apex:
spikelets solitary, densely imbricate in 2 series.
1.6-2 mm long. 0.8—1.2 mm
wide, villous, early deciduous, upper glume and
Spikelets ellipsoid,
lower lemma subequal. Upper glume membranous.
3-nerved, the margins rigid, densely pilose at the
basal portion with short and appressed, tuberculate
hairs, the margins ciliate, silky with hairs up to 1.5
mm long, the rest of the surface glabrescent. Lower
lemma glumiform, 3-nerved, pilose at the margins
and % of the surface. Upper anthecium obovoid.
1.5 mm long, 0.8-1 mm wide, plano-convex. mem-
branous, pale, shiny, rounded at the apex: upper
lemma smooth: lodicules 2. 0.1 mm long. condu-
plicate: stamens 3, anthers 1 mm long. Caryopsis
ellipsoid. 1 mm long: hilum punctiform, embryo
less than % the length of the caryopsis.
2n = 20 (Killeen, 1990).
Paspalum malmean-
Chromosome number.
Distribution and habitat.
um grows in Bolivia and Mato Grosso, Brazil, where
it is usually found in flooded savannas on sandy
soils between O and 500 m elevation.
Paspalum malmeanum is related to P. eucomum.
which differs by the spikelets 2.5—3 mm long. Oth-
er features of its distribution and habitat are point-
ed out in the treatment of P. eucomum.
Representative — LIVIA. Santa Cruz: Ve-
lasco. Parque Nacional Noel En M.. ju wise nto Los
Fierros, Gutiérrez et i? 55 (MO); Nuflo de Chavez, 3 km
SW of Concepción on road around reservoir. Killeen 2024
(G. LPB. MO. SI. US); Ñuflo de Chávez. Est. Las Madres
2076. (MO. SI, US); Ñuflo de Chávez, Estanci
O, SI.
fa. Killeen 2478 (M US).
f São Lourenço. E > F Ibo, Chase 11962
US): Juruena, Kuhlmann 7 (U =)
us Linha Telegrafica, на, p^ (1
— da
18. als Sage niquelandiae Filgueiras, Novon
: 30. fig. 1. 1995. TYPE: Brazil. Goiás: Mun.
Niquelandia, ca. 14?18'S, 48°23'W, ca. 2 km
este da localidad de Macedo. 19 May 1993, 7.
S. Filgueiras & F. C. de A. Oliveira 2401 (ho-
IBGE! isotypes, F not seen. G
K!, MEXU
sr,
lotype.
ICN not seen, not seen,
MO!, R not seen,
Figure 23.
seen.
ее SP see
nol seen, ` not seen,
US not seen).
Rhizomatous perennial; culms 120-185 em tall.
0.1—0.6 cm diam.,
elabrous. hollow: nodes glabrous, dark. Sheaths
unbranched; internodes 5-8.
glabrous or pilose toward the base and papillose-
hispid toward the distal portion. keeled, the mar-
Ligules 0.2-0.3 mm long.
membranous: pseudoligule a ring of hairs 5-8 mm
long. Leaf blades lanceolate, 20-45 cm long. 0.8-
gins ciliate or glabrous.
380
Annals of the
Missouri Botanical Garden
2.2 cm wide, acuminate, flat, glabrous to papillose-
hispid, the margins scabrous or ciliate. Peduncles
exserted, 12-22 cm long, glabrous. Terminal inflo-
rescences 14—18 em long, 6-9 ст wide; main axis
9-15 em long, glabrous, terete or wavy; racemes 8
to 30, 2.5-8.5 cm long, alternate, ascendent, end-
ing in a naked point, occasionally in a spikelet;
pulvini pilose, with 2-2.5 mm long hairs; rachis of
the racemes 1-1.5 mm wide, foliaceous, flexuous
toward the apex, glabrous or scabrous, midnerve
manifest and wings with thin, longitudinal nerves;
pedicels 0.1-0.3 mm long,
paired, arranged in 4 series. Spikelets ellipsoid to
scabrous; spikelets
ovoid, 2.3—3.3 mm long. 1-1.4 mm wide, pilose,
plano-convex. Upper glume membranous, 3-
nerved, the margins corky, with radiate, tuberculate
hairs 0.3-1 mm long, the rest of the surface gla-
brous or shortly hispid. Lower lemma flat, membra-
nous, J-nerved, glabrous, as long as the upper
glume, depressed the base. Upper anthecium
ovoid, 2.2-2.7 mm long,
nous to chartaceous, glabrous; upper lemma 5-
1-1.1 mm wide, membra-
nerved, with simple papillae regularly distributed
all over the surface and sparse bicellular micro-
hairs; lodicules 2, 0.1—0.2 mm long; stamens 3, an-
thers 1.2-2 mm long; stigmas plumose. Caryopsis
not seen.
Distribution and habitat. Endemic to the state
of Goiás, Brazil, where it grows on outcrops of ser-
pentine, rocky soils, near edges of gallery forests.
Paspalum niquelandiae is related to P. burmanii
and P humboldtianum. Differences with P bur-
manii were summarized in the treatment of that
species. In. Paspalum humboldtianum plants. are
(30—)50—100 cm tall, with branched or unbranched
culms, two to six racemes per inflorescence are pre-
sent, and spikelets are long ellipsoid, with radiate
hairs, 2-3.2 mm long, at the margins of the upper
glume.
19. Paspalum pectinatum Nees ex Trinius, Sp.
Gram. 1: pl. 1828. Paspalum pectinatum
Nees, Fl. Bras. Enum. Pl. 2 1829, hom.
illeg. Anastrophus pectinatus (Nees ex Trin.)
Schltdl. ex B. D. Jacks., Index Kew. 1: 118.
1893. TYPE: Brazil. Without locality, А Sellow
s.n. (holotype, LE not seen; isotypes, B!, G!,
K! photo, SI!, LE not seen, US-2942525!, WN.
Figures 18, 19.
Caespitose, shortly rhizomatous perennial; florif-
erous culms 35-100 ст tall, erect, few-noded; in-
ternodes 2 to 4, 7-14 cm long, 0.2 em diam., hol-
low, slightly compressed, striate, glabrous; nodes
brown.
glabrous, Sheaths usually longer than the
internodes, the basal ones shiny, red-tinged to
brown, glabrous; upper sheaths green, keeled, vil-
lous to hirsute all over the surface or only toward
the upper portion, the margins membranous. Lig-
ules membranous, 0.3-0.5 mm long. brown, gla-
brous; pseudoligule a tuft of white hairs 3-4 mm
10-30 cm long, 0.2-0.8
em wide, mostly basal, flat or with involute margins,
long. Leaf blades linear,
densely pilose on both surfaces, the base attenuate,
the apex acute. Peduncles long-exserted, up to 30
cm long, finely striate, glabrous. Terminal inflores-
cences with 2(3 to 5) conjugated, ascendent ra-
cemes, 2-8 cm long; main axis absent or up to 1
cm long; pulvini shortly pilose; rachis of the ra-
cemes shortly winged, 2-2.5 mm wide. glabrous,
membranous, green, with light-brown to purple,
crenate, glabrous or pilose, margins, the midnerve
manifest, ending in a naked point; pedicels up to
0.6 mm long, flat, glabrous; spikelets solitary,
densely imbricate and arranged in 2 series. Spike-
lets lanceolate, 5-8 mm long, 2-3.3 mm wide, dor-
siventrally compressed, slightly plano-convex, cor-
date, the apex acute, pilose, pale or tinged with
purple.
nerved, winged, papyraceous to hyaline, cordate,
Upper glume as long as the spikelet, 3-5-
slightly convex, glabrous, the base cordate, corru-
gated at the middle portion. Lower lemma 4.5-7.2
mm long, 3-nerved, not winged, sparsely pilose,
subcordate, the margins papillose-pilose with rigid,
tuberculate hairs 1-2.5 mm long. Upper anthecium
ellipsoid, 4-5 mm long, 1.4-1.6 mm wide, plano-
convex, papyraceous, pale; upper lemma finely pa-
pillose, with bicellular microhairs, short macrohairs
and isolated silica bodies at the distal portion and
along the margins; palea smooth: lodicules 2. ca.
0.4 mm long, hyaline, conduplicate: stamens 3, an-
thers 2 mm long, ferruginous; styles 2, free, stigma
plumose, lateral. Caryopsis not seen.
"Grama-das-pedras" (Brazil).
Chromosome number. 10 (Davidse & Pohl,
1978; Pohl, 1980); 2n — 40, 60 (Gould & Soder-
1967).
Distribution and habitat.
Common name.
strom,
This species grows
from southern Mexico, Mesoamerica, Colombia, Ve-
nezuela, Guyana, and Surinam to Bolivia, Paraguay,
and central and southern Brazil, mainly in savan-
nas on sandy soils with rocky outcrops. It is fre-
quent in areas with periodical fires, occasional in
margins of streams, between 180 and 1800 m ele-
vation,
Paspalum pectinatum is a member of section
Pectinata by its winged and cordate upper glume.
This species is related to P cordatum, a species
distinguished by its robust plants, 100—130 cm tall,
Volume 89, Number 3 Denham et al. 381
2002 Paspalum subg. Ceresia
Figure 18. Paspalum pectinatum.—A. Habit. —B. Detail of ligule. —C. Spikelet, ventral view. —D. Spikelet,
dorsal view. Е. Upper anthecium, dorsal view with lateral stigmatic plumules. —F. Upper anthecium, ventral view
with lateral stigmatic plumules. —G. Upper palea, lodicules, anthers, and gynoecium. (Based on Filgueiras 2459, Sl.)
382
Annals of the
Missouri Botanical Garden
MEN
1
|
¢ P pectinatum
e P phyllorhachis
ж Р petrense
— \ \
\ \
O 200 400 600 800 1000 km \ \
— ای یچ و \
° E 200 300 406 ме 600 т \ |
red bı Hendrik K "e \ \
— — —
Figure 19,
with 5 to 10 racemes per inflorescence, each ra-
сете 12-16 em long.
The basal portion of the plant is usually covered
with remains of burnt sheaths, which probably act
as a protection for the renewal buds
Phe specimens Fonseca et al. 153, Oliveira et al.
763, and Filgueiras 2459, collected in Goiás, Ni-
quelándia, Brazil, have 8 to 9 internodes with leaf
blades arranged along the floriferous culms.
There is a papyraceous bract at the base of the
inflorescence, as in Rivera 3, Pinto & Sastre 1066,
Rivera L-L41, Davidse 5321,
and
7649-۸ "
Cuatrecasas
Representative specimens.
Manatee (or Coastal) Hwy
Point,
BELIZE. Belize: along
y, 7.4 km S of Turnoff to Gales
Atha 46904 (MO). BOLIVIA. Beni: Prov.
Vaca Diez: Riberalta, 65 km hacia Santa Rosa, cerca del
desvío a Cobija, Beck 20552 (K). La Paz: Prov. Ren i
Luisita, Haase 630 (К, MO). Santa Cruz: Chiquitos, 4
m W of Santiago de Chiquitos, Killeen 2782a (MO, US
BRAZIL. Distrito Federal: Brasília, lago Sul, Zuloaga
3839 (SI). Goiás: Mun. Niquelándia, Macedo, ca. 20 km
Nee &
Distribution of Paspalum pectinatum, P. petrense, and P. phyllorhachis
de Niquelandia, dere 2455 (MO): Campo de Picho-
‚ Glaziou 2242 MO, P, US). Mato Grosso:
Expedition Base Ca un. — 6766 (К.Р, US): « ca. 70
km N of Xavantina, /rwin & Soderstrom 6758 (K, MO, SI)
Minas Gerais: Serra do Espinhaço, ca. 10 km SW o
iir ir Anderson et al. 35213 (MO); Diamantina,
Serra San Antonio, Chase 10350 (MO, US). Parana:
Ponta Grossi, — 104 (SI); Jaguariaiva, Dusén 16057
(MO, P, SI). ta Catarina: Campo Novo, Mafra, Klein
3916 (US). ом Paulo: Boa Vista, 4-5 km SE da Estação
Eng. Hermilo, Machado de Campos 111 (US). Without
— Sellow 1232, 1239 (B). Pe ea Arauca:
N del río Tame, Blydenstein 734 (COL). Boyacá: near
Ais ué, Haught 2709 (COL). Caquetá: Mun. San Vicente
del Caguán, Hato Caquetania, Bentacur & Porras 1510
(MO). Cauca: Chisquito, Asplund 10633 (US); Río —
йш, San José del Guaviare, Cuatrecasas 7649-A (С
IS). Cundinamarca: savanna of San Martín, SE of *
gotá, Shaw s.n. (US). Guaviare: Mun. San José del Gua-
лаге, Inspección de La Fuga. Giraldo-Canas & López
2631 (MO). Meta: 20 km SE of Villavicencio, Alston 7569
M. COL, US); Mun. Puerto Gaitán entre Carimagua y
El Porvenir, Rivera 3 (COL). Santander: La Laguna, car-
retera Lebrija-Pantano, Rivera L-141 (COL). Vichada:
Parque Nacional Natural Tuparro, Barbosa € Zarucchi
AA
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
3032 (MO); ca. 45 km F
between Las — and Santa Rita,
5321 (COL). COSTA RICA. Guanacaste:
E of Cumaribo along dirt road
Davidse & Llanos
Comelco, W of
Bagaces, Heithanus 163 (MO). San José: Cantón de
Acosta, Fila road Morales 4395 (MEXU. MO)
Puntarenas: Buenos Aires, Pohl & со 13922 (МО).
ТЕМА f j
a
H AS. — Haci че с Espíritu Santo, — 7445 (US).
MEXICO: Chiapas: Mun. ку nque, 2—4 km N of La Vic
NWN
toria and * km Palenque, Dandi 20544
(MEXU. MO). Oaxaca, — Sta. Marfa Chimalapa, El
Ocotal Grande, ca. 5 km al E de Sta. María, Hernández
2283 (MO). Tabasco: km 64 rumbo бе q rim
Rueda al O de Huimanguillo, Hill & Cowa 1548 (N
Veracruz: a 11 km « Choapas con la
carretera Cardenas-Coatzacoaleos, Orozco 182 (МЕХА).
NICARAGUA. Chontales: 15 km SE of Juigalpa, along
Hwy. 7. Pohl & Davidse 12361 (MO). Zelaya: near Cano
mba, — 5 km SE of El Empalme | ay road to Lim-
baika, Kral 69245 (MO). PANAMA. riqui: Cerro
Vaca, eastern C hiriquí, Pittier 4351 (US): C. erro Campana,
Lazor 3350 (MO). SURINAM. Sipaliwini savanna, Rom-
PARAGUAY. Canendiyú: Mbaracayú
tural Reserve ар red by Fundación Moisés Ber-
toni, 242115 'W. Zardini & Ramirez Benitez 51117
(MO). VENE "n E : i Amazonas: E of Parucito and
N of its tributary Caño Majagua, — 10153 (MO).
Apure: Distrito Pedro Camejo, 22 km W de Galeras de
Cinaruco, Davidse & González 15652 nn Bolívar: 2"
of La Vergareña. 5 of Ciudad Bolívar, Alba 53-23 (
Cano
20. Paspalum petrense A. C. Burman, Kew
Bull. 35: 297. 1980. TYPE: Brazil. Goiás: Ser-
ra de Е . 20 km E of Pire nópolis, 8
Apr. 1979, T. . Filgueiras & A. G. Burman
430 aud К; isotypes, IBGE! SP not
seen, RB-217458!). Figures 19, 20.
Caespitose perennial; culms erect, 70-90 cm
tall, unbranched: internodes terete, hollow, gla-
brous or sparsely pilose toward the summit; nodes
glabrous. Sheaths 7—10 cm long, papillose-pilose
toward the apex, overlapping, keeled. Ligules ca. 1
mm long, membranous, truncate; pseudoligule ab-
senl. 10-19 em
long, 3-6 mm wide, pubescent, the margins sca-
Blades linear to linear-lanceolate,
brous, attenuate at the base, the apex acute. Pe-
duncles 17-30 cm long, pilose at the apical por-
tion. Terminal inflorescences 8-15 em long, 4—7 cm
wide; main axis glabrous; pulvini pilose; racemes 3
to 5. 3-6 em long, alternate, ending in a naked
pines rachis of the racemes (6.5—)7—7.5 mm wide.
foliaceous, glabrous. green to purple. with anasto-
mosed veins on the wings, the margins slightly sca-
brous; pedicels 0.4—0.6 mm long. hispid, laterally
inserted in the spikelets; spikelets paired, the basal
one sometimes reduced or aborted, arranged in 2
› 4 series. Spikelets ellipsoid, 3-3.5 mm long. l-
A | mm wide, long-acuminate, pilose, pale, upper
glume and lower lemma subequal. with an annular
thickening at the base. Upper glume membranous
the
lemma membra-
to hyaline, 3-nerved, hairy on the lower half,
margins papillose-pilose. Lower
nous to hyaline, acuminate, 3-nerved, pilose on the
lower half with hairs shorter than those of the upper
glume. Upper anthecium ellipsoid, 2.8-3 mm long.
0.6-0.8 mm wide, membranous, shiny, with a tuft
of hairs at the apex, otherwise glabrous: upper lem-
ma and upper palea with bicellular microhairs and
papillae all over the surface and silica bodies at
the apex; lodicules 2, 0.1 mm long; stamens 3; stig-
ma plumose. Caryopsis not seen.
Distribution and habitat. Endemic to central
Brazil, Goiás, where it is present in rocky soils on
mountain slopes, at an elevation of 1200 m.
Paspalum petrense is related to P. heterotrichon
(see observations under this species) and P. trachy-
coleon. Paspalum petrense differs from P. trachy-
branching
coleon by its unbranched culms (vs.
culms), rachis (6.5—7.5 mm vs. 4-6 mm wide), the
upper glume long-acuminate (vs. acute to acumi-
nate), and lower lemma long-acuminate, pilose, not
suleate, and with inconspicuous nerves (vs. acute,
glabrous or with short, apical hairs, sulcate, and
with conspicuous nerves).
21. Paspalum phyllorhachis Hackel. Oesterr.
Bot. Z. 51: 240. 1901. TYPE: Minas
Gerais: without locality, 1872, A. К.М. Glaziou
20078 (holotype, W!; isotypes. B!. K! photo.
SI!, US-2855708! photo, SI!). Figures 19, 21
Brazil.
Shortly rhizomatous perennial: culms 70-170 em
tall, 2-7 mm diam., erect or climbing, bambusi-
form, branching at the lower and middle nodes: in-
ternodes 5-8 cm long, glabrous, terete, hollow. lig-
nodes glabrous, Sheaths 5-8
long, glabrous or pilose at the upper portion.
keeled,
membranous,
nified: brown. cm
margins smooth, glabrous. Ligules
1.5-2 mm long. truncate to rounded
ihe
—
at the apex, decurrent; pseudoligule absent. Leaf
blades linear-lanceolate, 7-17 em long. 0.4—1.3 cm
wide, rounded, long-acuminate at the apex, with the
abaxial surface glabrous to pilose and the adaxial
surface densely pilose, margins smooth to scabrous.
Peduncles exserted, glabrous or pilose toward the
distal portion. Terminal inflorescences exserted,
10-13 cm long,
5 mm diam., glabrous or pilose; racemes
1.5-3 em wide: main axis 4—9.5
em long, 0.5
7 to l - (:
a naked point; pulvini pilose; rachis of the racemes
5-)3.5-5 em long, alternate, ending in
1-6 mm wide, foliaceous, with anastomosed nerves
on the wings, glabrous, brown: pedicels 0.8 mm
laterally inserted in the spikelets:
long, hispid,
spikelets paired, imbricate in 4 series. Spikelets
384 Annals of the
Missouri Botanical Garden
— — —
— V
—
— —
Ja 7
de br y
MERE
Я
—
"fL xL Y — <
17 — — 3.4 =
— 3
=
1mm
AA
ERA
TAX
— т
—
ЖЧ
NN COM 9
Л A VÀ UNS
АШАА. Ч
ANA
M NW
^
D
'
V
\
Figure 20. Paspalum petrense.—A. Habit. —B. Detail of ligule. —C. Portion of the rachis. —D. Spikelet, dorsal
view, —E. Spikelet, ventral view. —F. Upper anthecium, dorsal view. —G. Upper anthecium, ventral view. Н. Jpper
palea and lodicules. (Based on Filgueiras & Burman 430.)
385
Denham et al.
Volume 89, Number 3
2002
Paspalum subg. Ceresia
D. Spikelet,
>, Upper anthecium, ventral view.
Figure 21.
dorsal view.
Portion of the rachis.
Paspalum phyllorhachis.—A. Habit. —B. Detail of ligule. С.
E. Spikelet, ventral view. —F. Upper anthecium, dorsal view. —(
H. Upper palea, lodicules, and anthers. (A based on Chase 8975, Us: B-H based on Glaziou 20070, Р.)
386
Annals of the
Missouri Botanical Garden
ellipsoid, 2.3-2.8 mm, 0.8-1 mm wide, glabrous,
upper glume and lower lemma subequal, scabrous
at the upper margins. Upper glume as long as the
acuminate,
nerved. Lower lemma glumiform, acuminate, mem-
hyaline to membranous, 3-
spikelet,
branous, 3-nerved. Upper anthecium lanceolate,
2.1-2.7 mm long. 0.7-1 mm wide, membranous,
scabrous at the apex, otherwise glabrous; upper
lemma and palea with conspicuous, simple papillae
and bicellular microhairs all over the surface, and
prickles at the apical portion; stamens 3, anthers
1.5 mm long; stigma plumose. Caryopsis not seen.
Distribution and habitat. Endemic to Brazil,
Minas Gerais, where it is found in open fields be-
tween 1200 and 1400 m elevation.
Paspalum phyllorhachis is related to P petrense
and P. trachycoleon by its paired spikelets. It differs
from both species by these spikelets being gla-
brous.
Glabrous and pilose spikelets were occasionally
observed in the specimen Magalhães Gomes 908,
with sparse hairs toward the apical margins of the
upper glume.
Representative — BRAZIL. Minas Gerais:
Serra do Curral, S « d Horizonte, Chase 9325 (US).
6975 (US) Rios d Pedras a Matto Grande, Glaziou
20070, 20076 (P); Belo Horizonte, Magalhães Gomes 908
(US-1762607).
22. се ева Nees ех Trinius,
. Paspalum polyphyl-
lum Nees, Fl. — Pl. 2: 41. 1829,
hom. illeg. TYPE: Brazil. Chapada, “in campis
siccis pr. 1824, С. Н. von Langs-
dorff s.n. (holotype, LE!; isotypes, K! photo,
SI!, US-2855764! photo, SI!). Figures 7, 22.
Gram. Panic. 4.
m
Mariannam,"
Paspalum —— ue m. & Schult. var. tenue
Doll, in Mart., Fl. Bras. 2(2): 67. 1877. TYPE:
zl. Without lac ality, L Riedel d (lectotype, —
nated here, W!; isolec totype, (
liri M Hack., бенен Во! 5
196. 1901 с 3razil. Rio de Janeiro: sion
locality, T oh rec — D 41 (holotype, W!;
type, US- с photo, SI!).
Paspalum bic T Mez, Re ре rt. Spec
15: 27. TYPE: Brazil. Goiás:
laco, casc se do Rio Macaco, 12 July 1895, А. E
M. Glaziou 22563 (lectotype, oa дем В!;
isolectotypes, BR not seen, K!, P! photo, SI!, US-
42158!
29421:
L
. Nov. Regni Veg.
Faze Pe do Sir-
Caespitose perennial, with long, arcuate rhi-
zomes, cataphylls coriaceous, villous; culms (10—)
25-90 cm tall, 0.1-0.2 em diam.. rigid, branched
or unbranched, erect, occasionally decumbent; in-
ternodes 1-10 cm long, terete, hollow, glabrous:
nodes brown, glabrous or shortly pilose. Sheaths
1.5-8 cm long, usually longer than the internodes,
glabrous to hirsute, with tuberculate hairs, the mar-
gins glabrous or shortly ciliate. Ligules 0.4—2.5 mm
long, membranous, brown, glabrous; pseudoligule,
when present, a tuft of rigid hairs 2.4—5 mm long;
collar glabrous to shortly pilose. Leaf blades linear
linear-lanceolate, 4-13 cm long, 0.1-0.6 cm
wide, flat, rigid, ascendent, glabrous to papillose-
pilose on both surfaces, narrow or rounded at the
base, the apex subulate, the margins short to long
papillose-pilose. Peduncles exserted, up to 30 cm
long, terete to filiform, glabrous to hirsute in the
distal portion. Terminal and axillary inflorescences
exserted; terminal inflorescences 4-12 cm long.
0.4—3(—4) cm wide; main axis absent or 1—4(—7) em
long, sulcate, glabrous to hirsute; pulvini shortly
pilose, with white hairs up to 0.6 mm long; axillary
inflorescences 4—7 cm long, 0.5-3 cm wide; ra-
cemes 1 to 4(8), ascendent, slightly divergent from
the axis, straight, alternate, ending in a nake
point, sometimes in a spikelet; rachis of the ra-
cemes 3-10 cm long, 1-1.2 mm wide, pilose or
with adaxial surface glabrous, green or tinged with
purple, with anastomosed veins, the margins sca-
brous and covered with long hairs; pedicels un-
equal, up to 1.7 mm long, triquetrous, hispid;
spikelets paired, densely arranged in 4 series,
sometimes lower spikelet aborted. Spikelets long-
1.4-3.8 mm long, 0.6-1.2 mm wide,
acute, pilose, the margins rigid, purple, covered
ellipsoid,
with long, ciliate spreading hairs; upper glume and
lower lemma as long as the spikelet, hyaline. Upper
glume with corky margins covered with radiate,
white hairs 2—7 mm long, alternating with shorter
hairs 0.8-2 mm long, otherwise shortly pilose, 3-
nerved. Lower lemma hyaline, the margins shortly
pilose, hairs up to 0.5 mm long, the rest of the
surface glabrous to sparsely pilose, convex and de-
pressed toward the base, 3-nerved. Upper anthe-
cium long-ellipsoid, 1.2-3.6 mm long, 0.6-0.8 mm
wide, acute, membranous, pale, with bicellular mi-
crohairs and simple papillae all over the surface
and macrohairs and silica bodies toward the apex
of the upper lemma and palea, upper lemma 3-
nerved; lodicules 2, ca. 0.2 mm long, conduplicate,
hyaline; stamens 3, anthers 0.8-2 mm long. Cary-
opsis long-obovoid, 0.8-1.3 mm long, 0.4—0.6 mm
wide; hilum linear, % the length of the caryopsis;
embryo Y to Y the length of the caryopsis.
Common name. — "Capim-lanoso" (Brazil).
Distribution and habitat. From Venezuela and
Colombia to central and southern Brazil, Bolivia,
Paraguay, northeastern Argentina, and Uruguay.
This species grows in open fields and savannas, on
Volume 89, Number 3 Denham et al. 387
2002 Paspalum subg. Ceresia
nu
; ЦР | VA
7
Ss
Don
f
Figure 22. — plan —A. Habit. —B. Detail of p E Portion of rachis. —D. Spikelet. =
view. —E. Lemma, ventral у Е Upper anthecium, dorsal view Upper — ventral view. —
Caryopsis, scutellar view. —I. С ‘aryopsis, hilum view. (Based on Smith y — 12106, US.)
Annals of the
Missouri Botanical Garden
rocky, sandy soils, occasionally in swamps, between
О and 2500 m.
Paspalum polyphyllum shows a high degree of
plasticity in its vegetative characters, such as plant
height, branching, blade pilosity, and size of inflo-
rescences and spikelets. This species has two dif-
ferent growing patterns; one of them, with simple
culms bearing larger inflorescences and spikelets
of different size, from 2.6 to 3.6 mm long (e.g..
Swallen 8772, Smith & Klein. 12106, Hassler
11561, Schulz 17275, Archer & Barreto 4995, and
Fiebrig 6222) more than 4
13933, which also have inflorescences more than
mm, e.g.. Rojas
19 em long, racemes 11-13 cm long, and spikelets
4—4.2 mm long. The other phase is characterized
by its culms profusely branching at the middle
nodes, with smaller inflorescences and spikelets ca.
2 mm long (as in /rwin et al. 16232a, 17256, Kil-
leen 2795, Hunt & Ferreira Ramos 6012, Hunt
6246, and Chase 9148). Also, both phases can be
Swallen 9150,
where simple, erect culms up to 40 cm tall, with
found in a single collection, e.g.,
spikelets 3 mm long, are mixed with more delicate
culms in which spikelets are 2 mm long. It is note-
worthy to mention that the type material of P. po-
lyphyllum belongs to the second phase of devel-
opment.
When studying type material of Paspalum bici-
lium and P. macroblepharum it became evident that
they belong to the branching phase of P. polyphyl-
lum, with spikelets up to 2 mm long.
Doell (1877) cited, following the concept of Tri-
nius (1828-1836, figs. 134, 144). P. ble-
pharophorum and P. distichophyllum what here is
under
considered P. polyphyllum, including in P. blephar-
ophorum specimens of the unbranching phase and
in P. distichophyllum specimens of the branching
one. Both species are here considered synonyms of
P. humboldtianum (see synonymy under this spe-
cies).
Doell (1877) cited three syntypes when describ-
ing P. blepharophorum var. tenue, of which Riedel
s.n. was here selected as lectotype of the species;
of the remaining syntypes, Sellow s.n. also belongs
to P. polyphyllum, and D'Orbigny 168, from Chu-
quisaca, Bolivia, probably belongs to P. humbold-
tianum,
The number of racemes is higher, ranging from
4 to 12, in the specimen Chase 9369.
Mez (1917) cited two syntypes when describing
P. bicilium, Glaziou 22562 and 22563, of which the
latter has been selected as lectotype of the species.
According to Rosengurtt et al. (1970) this spe-
cies is a low-vield forage grass.
Representative specimens. ARGENTINA. Chaco:
Dpto. San Fernando, Fontana, Meyer 2090 (LIL); Dpto. 1°
de Mayo, Colonia Benítez, * 17275 (G, MO, SI).
Corrientes: Dpto. Paso de Los Libres, Campo Militar
. Carnevali 3200 (CTES, LIL). Misiones:
Dpto. San Ignacio, sobre Ruta Prov. 210 m, a 7 km Ayo.
Horqueta, Honfi 98 (LIL, MO); Puerto Nuevo, Schwarz
2214 (B, LIL, US). BOLIVIA. La Paz: Puente Villa-Yun-
gas, Cárdenas 3606 (US). Santa Cruz: Prov. Chiquitos, S
slope of the M d _ Santiago, Daly et al. 2171 (MO,
) iqui ia de Santiago, Killeen 2 (MO);
Buena Vista, St 6 (BM, x K, MO, W, US).
BRAZIL. Bahia: Mun. Palmeiras, Pai Inácio, e 242,
km 232, Mori 13303 (К. MO); Dpto. Palmeiras, Morro do
Pai Inacio, Zuloaga & Morrone 6944 (IBGE, SI). Distrito
Federal: a 32 km do CENARGEM, Borges Dias Vieira
189 (MO); Parque icipal do Gama, ca. 20 km S of
Brasília, Irwin & Soderstrom P (MO, SI, US). Goiás:
Serra do Caiapó, ca. 33 S of Caiaponia on road to
Jataí, Irwin & Sode "rstrom 7069 (К. MO, SI); between Jataf
and C aiaponia. 45 km fre КО иу нсә Нит 6246 (К, US);
Serra Azul, about 7 km W of the Barra do Garcas, —
& Ramos 6012 (К, US). — Grosso: ca. 6 km S of
Xavantina, Argent et al. 6460 (P, US); Serra do Roncador,
tal. 162: Za (MO); summit, Serra Azul,
—
Glaziou 20082 (P, US, W); Serra do Cipó, Chapeu de Sol,
Santa Luzia, ris & Barreto 4995 (US); Serra
110 km NE of Belo Horizonte, Chase 9148 (M0);
Preto, Villa Rica, Chase 9369 (MO, US). Paraná: I:
NE of Atuba (which is n outside of — along Hwy.
‚ Davidse et al. 10967 (MO); Po ^R Gros "azenda
de Chacon, Swallen 8772 (US). Rio le Janeiro: а
rade & Vianna 20377 (US); environs de Rio de Jane
Glaziou 17940 (G. MO, W). Rio Grande do Sul: Mo n-
tenegro, Araujo 144 (US); Pelotas, Swallen 9150 (US).
Santa Catarina: Mun. Mafra, campo 10 km NW
on the road to Barrac ‘as (20 km), Smith & Klein 12106
ж ) Sao Pauli
=
~
—
íchica, Sarai 4186 (COL). Santander:
, Robinson « Beltran 3189 (US). PAR-
AC SU, AY. Alto Parana: pi re — ои Alto Р агапа,
Fiebrig 6222 (BAA, BM, US). Ama — in
campis arenosis Estrella, — nod (G, US).
uaz: Caaguazú, Balansa 90a (G, P). Cer iral: E re-
gione lacus Ypacaray, Hassler 11. 501 (MO). Cordillera:
ед Colonia Pedro Caballero, Rojas 13404 (LIL-
61328). Guairá: Borja, Montes s.n. (LIL-561328). Par-
i: Cerro Peron, prés de deca Balansa 90 (BM,
G, P. US, W). URUGUAY. Art —
(BAA, W). Cerro Largo: — Herter
US). Flores: Río Yí y Ayo, Matanzas, — s.n. a
1723957). Paysandú: Río Uruguay y Chapicuy, Rosen-
— + n. (LIL, US). Rivera: Tranqueras, Castellanos
17922 (LIL). Tacuarembó: without locality, Osten 6632
(SI, ic 5). VENE A. Lara: Distr. Carretera de
Humocaro Bajo via Buenos Aires, 2 er W erff & Rivero
7863 (MO). Trujillo: | a C riste ur — 1867 (US).
23.
— pa )
~
Paspalum reticulinerve Renvoize, Kew
Bull. 50: 339, fig. 1 A-G. 1995. TYPE: Boliv-
ia. La Paz: Iturralde, Siete Cielos, Río
Manupare, ca. 2.5 km al E, 12°27’S, 67?37' W,
180 m, 8 June 1987, J. C. Solomon 17003
(holotype, LPB not seen; isotypes, K!, MON.
Figure 1
Prov.
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
389
Shortly rhizomatous perennial; floriferous culms
(70—)90—120(—140) ст tall, 0.3 em diam..
branched or unbranched:
erect.
internodes finely striate,
glabrous, hollow; nodes 5 to 7, compressed, brown,
glabrous. Sheaths 15-24 cm long, usually longer
than the internodes, keeled toward the upper por-
the inner surface red-
tion, glabrous to hirsute.
tinged-brown, shiny, the margins membranous, gla-
brous. Ligules membranous, 1-2.5 mm long.
brown, glabrous: pseudoligule absent or a tuft of
17-
40 em long. 0.3-0.8 em wide, flat, glabrous to hir-
hairs 5-7 mm long. Blades linear-lanceolate,
зше, the base attenuate and the apex acute, the
margins involute, with long hairs toward the lower
portion, otherwise scaberulous. Peduncles up to 40
em long, terete, glabrous. Terminal inflorescences
exserted, 5—15 em long: main axis absent or up to
l em long, flat, glabrous: pulvini shortly pilose: ra-
cemes (1)2(3). i
point;
naked
1-1.8 mm
glabrous 1
5-15 cm long. ending in ¢
rachis of the racemes winged,
wide, green or tinged with purple,
)
sparsely pilose, the midnerve scabrous and margins
scaberulous to papillose-pilose; pedicels unequal,
flat. scaberulous: spikelets paired, densely imbri-
cate in 4 series. Spikelets broadly ovoid, 4.8-6 mm
long. 3—4 mm wide, dorsiventrally compressed, pla-
no-convex, winged, cordate, ciliate on the margins
of the upper glume. Upper glume as long as the
spikelet, 7(—9)-nerved, winged, with just 3 central
nerves reaching the apex, membranous, ciliate, cor-
date, the apex acute. Lower lemma 3.5—1.4 mm
long, 2 mm wide, 3-nerved, glabrous, coriaceous,
corrugated and rounded at the base, the margins
winged, shortly ciliate. Upper anthecium ellipsoid,
2.8-3.6 mm long, 0.8-1.5 mm wide, plano-convex.
papyraceous, pale; upper lemma with simple pa-
pillae, bicellular microhairs, and macrohairs toward
the apical portion; upper palea with papillae and
bicellular microhairs; lodicules 2, ca. 0.4 mm long,
hyaline, conduplicate; stamens 3, anthers 1.6 mm
long; styles 2. stigma plumose, lateral. Caryopsis
obovoid, 1.6 mm long, 0.8 mm wide: hilum elliptic.
embryo Y the length of the caryopsis.
Iconography. Renvoize, Gramíneas de Bolivia:
450, fig. 97
Distribution and habitat. Present in Bolivia, in
La Paz and Santa Cruz, and Brazil. in the states of
Pará, Tocantins, and Mato Grosso. It grows in open
fields and “campos rupestres," between 100 and
900 m.
Paspalum reticulinerve is a member of section
Pectinata, where it is related to P. aspidiotes and
the latter
spikelets and inflorescences with (2)5
solitary
to 7(11) ra-
Р. imbricatum; has glabrous,
cemes. When describing P. reticulinerve, Renvoize
(1995) related this species to P. aspidiotes, from
which this author distinguished P reticulinerve by
the densely pubescent, attenuate leaf blades, a cil-
iate rachis, and smaller spikelets. Later, Renvoize
(1998) segregated Р. aspidiotes from P. reticulinerve
by having glabrous and cordiform leaf blades, the
rachis scabrous, and a smooth upper anthecium.
The study of type material of both species, and
specimens from Bolivia and Brazil, allowed us to
conclude that the real identity of P aspidiotes had
been misunderstood: this taxon is clearly separated
from P. reticulinerve by having solitary spikelets
6.5—7.5 mm long. Also we concluded that the pi-
losities of leaf blades and inflorescences are unre-
liable characters to segregate both species: in the
specimen Kuhlmann 1671 the rachis of the ra-
cemes is sparsely pilose, with hairs up to 0.4 mm
long, while in Killeen 6541 pilose and attenuated
leaf blades are present together with a glabrous ra-
chis of the racemes. A similar situation is present
in the specimen Plowman et al. 8949, with the ra-
chis of the racemes sparsely pilose.
Representative specimens. BOLIVIA. Santa Cruz
Prov. Velasco, Parque Nacional Noel Kempff M., Serranías
de Caparuch, Killeen 6541 (MO); жа Nacional Noel
kempff M.. Killeen et al. 4824 (SI). BRAZIL. Mato Gros-
so: Serra y d estrada do Di: amantina, Kuhlmann
1671 (US). rabá, Serra dos Carajás, Cavale —
2131 (US): e fone eicáo do Araguaia, about 4 km V
r along hwy. PA-150 pb
(MO, US). Tocantins: Mun. Сошо M: — 5; Maus iro
along GO-70, Plowman et al. 9133 (MO, US
of town cente Plowman et al.
24. Paspalum stellatum Humboldt & Bonpland
ex Flüggé. Gram. Monogr., Paspalum 62
1810, as “Paspalus stellatus.” TYPE: Colom-
bia. Without locality, А W. H. A. ron Humboldt
A. J. A. Bonpland s.n. (holotype, B-W not
BM!, US-80051!). Figure 23.
&
seen; —
TE stellatum Humb. & p ex Flügg ke mon-
/ Fl. Br РІ. 2: 78. 1829, as
a " TYPE:
achyum Nees, ras. Enum.
I
5 2
'aspalus stellatus var. — һу
zil. Without locality, £ Sellow 2 esp е. M not
seen; isotypes, B not seen, P!, 052! photo, 511).
Paspalum ste — — oe ex и var. dis-
. Enum. PL 2: 78. 1829, as
TYPE: Brazil.
G. H. von Langs-
— ме pi var. ye ds”
“In apricis arenosis prope — x
dorff s.n. (holotype, LE not se
Gram.
=
Paspalum cujabense Trin., Sp.
3: pl. 284. 1829—
1830. TYPE: Brazil. s Grosso: Cuiabá, in gra-
minosis — 1829 Н. von Langsdorff s.n.
(holotype, LE! isotypes, - photo, SI!, US-80050!).
m o Schltdl., Linnaea 26: 13. 133
854. TYPE: zuela. Distrito Federal: Surface-
. Wagener . 396 tà ype not seen).
Paspalum ас Нас К. var. sphacelatum Hack., Oes-
51: 239. 1901. TYPE: Brazil. Goiás:
390
Annals of the
Missouri Botanical Garden
ж P malmeanum
У Р niquelandiae
W Р stellatum |
e Г longiaristatum |
|
km \ 1000 800 600 400 200 0
\ لہ مایم کا ق |
Шы 200 300 400 500 600 miles | °
d bs Hendrik М Ks phem \
Figure 23.
Santa Luzia, 8 Abr.
1895, A. КОМ. — т
— W!: US
isoly ypes P!, US-2856
ene ш. Ий, & le ex ¢ Flüggé f. hirsuta
ack., in Stuck., Anales Mus. Nac. Buenos Aires
. TYPE: Argentina. Chaco: Resistenci ча,
‚ Н. Mercenaro s.n., leg. T. J. a
4 holot type W!: isotypes, CTES!, SI! |
a photo, SI!).
Caespitose, shortly rhizomatous perennial, cata-
phylls pilose; culms 25-100 ст tall, 0.1—0.2 em
diam., unbranched, erect to geniculate, fasciculate
at the base; internodes 3-18 cm long, terete, hol-
low, glabrous: nodes 3—6, brown, glabrous to shortly
pilose. Sheaths 4—8 cm long, striate, glabrous to
hirsute, with tuberculate, deciduous hairs, the mar-
gins membranous; basal sheaths overlapping. Lig-
ules membranous, short, 0.2 mm long, brown, gla-
brous; pseudoligule a tuft of short hairs intermixed
with long hairs up to 0.5 mm long; collar glabrous.
Leaf blades filiform, 8-27 cm long, 0.1-0.3 cm
wide, mainly basal, ascendent,
rigid, narrowed at
Distribution of Paspalum longiaristatum, P. malmeanum, P. niquelandiae, and P. stellatum.
as the spikelet.
the base, apex acuminate, the margins involute, the
basal ones ciliate, both surfaces sparsely papillose-
hirsute covered with long, tuberculate hairs, or the
adaxial surface glabrous. Peduncles up to 25 em
long, glabrous to shortly hirsute. Terminal inflores-
cences exserted; main axis absent; pulvini shortly
pilose, with (1—)2 papyraceous bracts, up to 4 mm
long, at the base; racemes 1, rarely 2 conjugated,
falcate, ending in a naked point; rachis of the ra-
cemes broadly winged, 2-15 ст long, (4—)5—10 mm
wide, glabrous to shortly hispid on the adaxial sur-
ace, green to purple, midnerve prominent, the mar-
gins membranous to hyaline, nerveless, brown, flat
to plicate and enclosing the spikelets; pedicels 0.1—
0.2 mm long, shortly hirsute, with radiate hairs at
the apex: spikelets solitary, arranged in 2 series.
Spikelets ovoid to ellipsoid, 2.4—3.9 mm long. 1.2—
1.6 mm wide, plano-convex, acute, pale or tinged
with purple toward the apex or all over the surface,
villous, the upper glume and lower lemma as long
Upper glume membranous, 2-3-
—
Volume 89, Number 3
Denham et al.
Paspalum subg. Ceresia
nerved, the central nerves inconspicuous or absent,
with a tuft of hairs toward the base, glabrous on the
rest of the surface, the margins papillose-pilose.
with white hairs up to 4 mm long. Lower lemma
glumiform, 2-nerved, the nerves marginal. Upper
anthecium narrowly obovoid, 1.8-2.9 mm long,
0.8—1 mm wide, 0.4—1 mm shorter than the upper
glume and lower membranous, pale.
smooth, glabrous, shortly stipitate, easily deciduous
at maturity; upper lemma 5-nerved, smooth, with
lemma,
sparse bicellular microhairs; lodicules 2, 0.2 mm
long: stamens 3, anthers 1.8-2 mm long. Caryopsis
obovoid, 1.6 mm long, 0.8 mm wide; hilum elliptic;
embryo Y the length of the caryopsis or slightly
shorter.
U.S. Natl. Herb.
Las Gra-
Iconography. Chase, Contr.
28(1): 16, fig. 1. 1929.
mineas del Distrito Federal:
Burkart, Fl. Il.
Rosengurtt et al.,
Luces de Febres,
115, 65. 1963.
‚ fig. 156. 1909.
Gramíneas uruguayas: 348, fig
151. 1970. Pohl. Fieldiana, Bot. 4: 454, »
171. 1980. Smith et al., Fl. Il. 922,
fig. 185. 1982. Rodríguez, Ernstia 8(2—3): 46.
1998. Renvoize, Gramíneas de Bolivia: 452, fig. 98.
D. 199
— name.
fig.
Entre Ríos: 377
n.s.
Catarinense:
“Capim-estrela” (Brazil).
Chromosome number. 2n = 32,52 (Honfi et al..
1991[1990]); 2n = 20, 32 (Killeen, 1990).
Distribution and habitat. From southern Mexi-
co, Central America, and the Caribbean to South
America, where it is found in Venezuela, Colombia.
Bolivia. Brazil. Paraguay, and Argentina: cited for
Uruguay by Rosengurtt et al. (1970). It grows in
fields or savannas on sandy or rocky soils, or lat-
eritic outcrops, between 0 and 2000 m elevation.
Paspalum stellatum is related to Paspalum cer-
esia, P. eucomum, and P. malmeanum by its winged
and nerveless rachis of the racemes, with spikelets
densely pilose. Paspalum ceresia differs by its glau-
cous plants, with filiform to linear-lanceolate leaf
blades, 7-21 em long, 0.3-1.522) em wide: inflo-
rescences with two to seven racemes (rarely a sol-
itary raceme), alternate, without a bract at the basal
pulvini; and pedicels hispidulous and upper anthe-
cium not stipitate. Paspalum eucomum is distin-
guished by having two conjugated racemes (rarely
a third one) per inflorescence, rachis of the racemes
1.5-2.5(—4) mm wide, and upper anthecium as long
as the upper glume and lower lemma. Paspalum
malmeanum differs by its rachis of the racemes
1.2-2 mm wide, spikelets 1.6-2 mm long. and up-
per anthecium as long as the spikelet.
Compound papillae are occasionally present in
the upper anthecium of specimens from Santa Cruz,
Bolivia.
According to Rosengurtt et al. (1970) and Smith
et al. (1982), this species 15 a low-yield forage
orass
grass.
Representative specimens. ARGENTINA. Chaco
Dpto. 1° de Mayo, Colonia Benítez, Schulz 300 (LIL). Cor-
. Goya, Paraje San Isidro, 46 km S de Goya,
1 E TES, K, LIL). Entre Rios: Dpto.
, barrancas del Río Uruguay, Bur-
(SI). Formosa: — we with-
y. Joergensen n 2882 (C ‚ MO, SI, . Misiones:
ija, Ekman МУ (US); Dpto.
a, Montes 1953 3 (G, LIL, P. US, W).
Beni: Prov. Iténez, Magdalena, Moraes et al.
1713 (L PB, SI). La Paz: Chaquimayo-A polo trail ca. 15
y & Foster 71186 (MO). Santa
O
rientes: Dpto
Ahumada etal. 354
—
Cruz: Pr dase San Ignacio, Beck & Seidel 123
(MO. SI); Nuflo de Chávez, Concepción, Killeen 2474
(MO, SI, US). BRAZIL. Amazonas: Uraricuera. Myers
3404 (US). Distrito Federal: Brasilia, campus da UNB,
de Jesus 16 (US). Goiás: Rio Cristal, 44 km by road SE
of Cristalina, Anderson E (ВМ. MO, US): without lo-
cality, Glaziou 22546 (G, K, P, W). Maranhao: Mun. Lor-
eto, Ilha de Balsas region, * & Ейеп 4417 (K, US).
Mato Grosso: Mun. Terenos, Faz. Modelo (IPEAO). Al-
MO); ca. 270 km N of Xavantina,
Ratter уз en 1915 (К, US). Minas Gerais: Uberaba. Chase
), US). Parana: Jaguariaiva. Dusén 18006 (BM.
No Es Rio Grande do Sul: São Gabriel, Araujo
(US). Roraima: Campos Gerais ahi Regiáo de €
x. Fidei 23626 (US). Santa Catarina: Chapecó. 9
km W of Campo Ет, Smith & Klein 11529 (US). Sao
campos d'Ipanema, — 13334 (P, US). CO-
. Arauca: 13 km al 5 de Arauca, laguna El Ve-
pe rgensen 52 (C OL). үн ен Cerro San. Lucas.
uie E-747 (US) »yaeá: Llanos ш ntales, El Yopal
al S del Hato Matapantano, Blydenstein & Saravia 1186
(COL). Magdalena: Santa Marta, Smith 142 (BM, K, MO,
US). Meta: Llanos Orientales, — de la selva de Arau-
ca, Blydenstein 736 (COL). e de Santander: Los
Estoraques, La Playa, Balick (C OL). COSTA RICA:
MA : km » Carretera Interamericana
along the road Boruea, . 11594 (US
MINIC AN RE PU BL IC. Dajabón: Cordillera Central, 6
km de Partido al W en la carretera a Dajabón, Zanoni et
Cara-
oen
LOME
al. 31952 (US). Monte Cristi: Sabaneta, Leonor, Valeur
198 (MO, US). HAITI. Circa Carrajal, Holdridge 1858
(MO, US). du Nord: vicinity of = Michel de PAtalave
Leonard 7537a (US). EL SALVADOR. San —
Volcán de San Salvador, Calderdn 2272 (US). JATE-
MALA. Jalapa: between Monjos and — a 10 mi.
S of Jalapa. Steyermark 32212 (US). HO Co-
E —
Francisco
Шеше, near
mayagua: Siguatepeque, Clewell 3514 MO).
road to are Swallen EE F (MEXU, US).
km by road SE of Rio Agua Cz
‚ Pohl € a do (MO). MEXICO.
Chiapas: near Caucua, Nelso (US) Mun. Villal
Corzo, above Colonia Vincente c uerrero on road to Finca
Cuxtepec, Breedlove & — 54586 (MEXU). Oaxaca:
Distr. Inquila, Conzali 4364 (MEXU, US). PANAMA. F
clé: between Paso de eh and Ola, Pittier 5020 (US
Panama: ca. 10 km SW of San Carlos along the ns T-
American Hwy., Davidse & D'Arcy 10126 (MO). Vera-
guas: on Santiago—Santa Fe rd., 18 km S of Santa Fe, Nee
Annals of the
Missouri Botanical Garden
8166 (МО). PARAGUAY. Alto Paraguay: Fuerte Olim-
po, Anisits 2047 (Р). Amambay: — y al sur de Bel-
avista, Fernández Casas € Molero 6335 (G); in regione
calcarea cursus * rioris — pes Hassler 11058
(BM, ( ч Р, US, W). Caaguazú: С aaguazú, Hassler 9259
, LIL, “MO > W. "Ca aazapa: Tavaí, Basualdo
2162 — MO, SI). € entral: ' larumandy, Schinini 6207
(G, LIL, SI, US). Cordillera: Barrerito, Anderson 1120
(US). Guairá: Villa Rica, Joergensen 4568 (BM, MO, SI.
US). Itapúa: gr — 6 km de la Ea. Melgarejo.
Quintum et al. SI). Misiones: San Ignacio, Burkart
Paraguarí: Cerro Peron, prés de Paraguarí.
— — G, K, P, US); C ordillera de Altos, Fiebrig
1 (BM, G, K, P, SI, US, W). VENEZUELA. Aragua:
М stween S. don de los — and U berito, Pittier 11326
I5). Barinas: н km SW of the Mé He intersection just
E of Barinas, Davidse 3165 (MO). Bolivar:
Divina Pastora, Timaya 2881 (US). Carabobo: V
— near Guacara, Pittier — (US
km W of San Carlos along Hwy.
Migne Federal: Caracas and vicinity, Bailey & Bailes
2 (US). Lara: Dpto. Palavecino, between the quebrada
La а “Mata and the E side of the ы А Nacional Terepaima,
Burand Jr. & а b > (MO). Miranda: Colonia
Tovar, Escalona et al. 1 (MO). Portuguesa: Dpto.
Guanare, terrenos de la Ho Ramírez 1789 (MO).
18242 (SD.
25. Paspalum trachycoleon Steudel, Syn. РІ.
Glumac. 1: 28. 1855 [1853]. TYPE: Venezue-
la. Valencia, 9 Oct. 1843, Н. C. Funck 742
(holotype, P!; isotypes, BM!, K!, G!, P!, US-
80059!, W!). Figures 24, 25.
Shortly rhizomatous perennial; culms erect to de-
cumbent, (0.8—)1—2 m tall, 0.3-0.6 mm diam.,
branching at the middle and upper nodes; inter-
nodes 5-15 cm long, terete, glabrous, rigid, hollow;
nodes pilose to glabrous, brown. Sheaths usually
longer than the internodes, overlapping, pubescent
to papillose-pilose at the upper portion. Ligules 2—
3 mm long, membranous, glabrous; pseudoligule
absent. 10-20 en
long, 0.5-1.5 em wide, flattened, pubescent to pa-
Leaf blades linear-lanceolate,
pillose-pilose on both surfaces, otherwise glabrous,
the base attenuate and the apex acute. Peduncles
terminal, partially exserted, pilose. Terminal inflo-
rescences 5-17 cm long, 3-5 cm wide; main axis
3-11 cm long, pilose; racemes 4 to 12, ascendent,
divergent, alternate, ending in a naked point; pul-
vini pilose; rachis of the racemes 3-10 cm long, 4—
6 mm wide, foliaceous, green to purple, with a pu-
bescent keel and anastomosed veins on the wings,
glabrous, ending in a naked point; pedicels un-
equal, hispid, laterally inserted in the spikelets;
spikelets paired, densely arranged in series.
Spikelets ellipsoid, 2.3-3.5 mm long, 1-1.1 mm
wide, pale, pilose, with an annular thickening at
the base. Upper glume as long as the spikelet,
the
margins pilose with white hairs, a few of them up
acute to acuminate, hyaline to membranous,
to 3 mm long, densely pilose toward the base, oth-
erwise glabrous, 3-nerved. Lower lemma as long as
the upper glume, acute, membranous, sulcate and
with conspicuous nerves, 3-nerved, pilose at the
upper portion, otherwise glabrous. Upper anthe-
cium ellipsoid, 2-3.1 mm long, 0.8 mm wide, pla-
no-convex, membranous; upper lemma with simple
papillae and bicellular microhairs all over the sur-
face and а tuft of apical macrohairs; upper palea
with simple papillae and bicellular microhairs to-
ward the distal portion; lodicules 2, reduced; sta-
mens 3, anthers 1.3—1.4 mm long; stigma plumose.
Caryopsis not seen.
Chromosome number. = 20 (Davidse & Pohl,
1974).
Distribution and habitat.
co, Guatemala, Honduras, and El Salvador to Ve-
nezuela, Colombia, and central and southern Brazil.
From southern Mexi-
It grows in mountain savannas, on dry mountain
slopes, between 500 and 2000 m.
Paspalum trachycoleon is close to P. phyllor-
hachis in its similar habit and inflorescence type,
which is distinguished by its glabrous spikelets. It
is also related to P. heterotrichon, but the latter has
the rachis of the racemes 3.54 mm wide and
spikelets solitary.
The specimen Killip & Smith 19038 has spike-
lets 3.8-4 mm long.
Representative — BRAZIL. Distrito Feder-
al: Reserva Ecológica do IBGE, da Silva 2659 (MO); ipa
Ribei irão Torto, NE of Lagoa Paranoá, Irwin et al. 15.
| Goias: Niquelándia, Filgueiras 3567 (SD; sed
E * — Burchell 7061-2 (K), Gardner 4390 (BM, G,
К, Minas Gerais: without locality, Glaziou 20079 (К).
loc — without
Cordil-
ЕШМШ, Cundinamarea: without
collector 80 (US-1574261). Huila: este de Neiva,
lera Oriental, Rusby & Pennell 1014 (US).
Estoraques, La Playa, Balick T (C
icini › & Smith
; vicinity of Charta, Killip & Smith 19038 (US).
EL SAI VADOR. Copán: 3 km NW of Copán, Williams et
al. 42957 (US). San Salvador: Volcans of San Salvador,
Hitchcock 8946 (US). GUATEMALA. Zacapa: slopes of
qee Virgen, Steyermark 42586 (US). HONDURAS. El
m — to Yusc aran. Swallen 11353
А со Morazan: Rio de Gallo,
near El Jicarito, Swallen. 10993 (MEXU. MO, US). Olan-
cho: Jutiapa — Camp, Pohl & Gabel 13761 (MO).
XICO. Chiapas: Mun. Cintalapa, 23 km of Las
Cruces — ey to La Mina Microwave Ballon; Breed-
love & Davidse s.n. (MEXU-830597); Mun. Villa Corzo,
above б, Vincente Guerrero on road to Finca Cux-
Breedlove & Davidse 54564 (MEXU). VENEZUE-
nzoategui: Distr. Libertad, road from El Vigia to
Bue nos Aires, Davidse & González 19487 (MO). Aragua:
de Maracay a Alto, Croat 21469 (MO). Distrito Federal:
Caracas, Bailey & Bailey 447 (US).
Jimenez, e ntre — del Viento y Cerro Pando, Davidse &
González 21162 (SI). Miranda: Sebastopol, Badilla 100
tander: Los
lepec,
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
Portion of rachis. —D. Spikelet, dorsal
'pper anthecium, ventral view. Н.
amayo
ycoleon.—A. Habit. —B. Detail of ligule. С.
i iew. —G. 1
lse & Tillet 4514, MO; В—1 based on 7
ispalum trach
ventral view. =. Upper anthecium. dorsal
Caryopsis, scutellar view. —L Caryopsis, hilum view. (A based on Dari
let
Figure 24. Pe
view. —E. Spike
1084, US.)
394
Annals of the
Missouri Botanical Garden
|
Ф P imbricatum |
e Р trachycoleon |
|
\
à 200 400 600 800 1000 km \ \
— — —
о ады 200 300 400 500 600 m: \
pared by Hendrik R. K
DERE. Es 7-96
— —
Figure 25.
5 —
Distribution of Paspalum imbricatum and
(US); San н de Morros, Davidse & Tillet — (MO).
Mona : S-faci oe mountain slope direc
Piele et gu ithout depar
cas, Tamayo d MES Los Venados, amaro. 2046 (US).
EXCLUDED SPECIES
Paspalum sanguinolentum Trin. was placed in
the subgenus Ceresia by Chase (1929), but this spe-
cies belongs to subgenus Paspalum, group Erian-
tha.
Literature Cited
Arriaga, М. О. 1987. Interpretación rA ornamento de
antec io de pm ‘hloa (Poaceae). Bol.
: 13 xs
Beetle; A. — Noteworthy grasses from Mexico X.
ai 52: 11-17
. Argent. Bot.
Paspalum, Notes on Gramineae. J.
‚19: 34—3t
— h support and tree stability. Cla-
Brooks. R. R. ps m ntine and Its Vegetation. Dios-
corides Press, Portland.
Р,
trachycoleon.
Burkart, A. 1969. Gramíneas. /n: A. ir —
ra de Entre
. Agropecu. 6(2a): 1
Canto-Dorow, T. S.,
Flo-
. Nac. Tec-
—
ч i ;olecc.
pm Wagner & J. F. Monte-
negro Valls. 1996. Re visão taxonómica das espécies de
Paspalum L. grupo Notata (P o i Rio
— do Sul, Brasil. Iheringia, ic 17:
Chase l In: tcheoc A "s
— ue bn Peru, and — a U.S. Natl.
Herb. 24: 434—455.
. 1929. The North * rican = ies of Paspalum.
U.S. Natl. Herb. 28: 1—31(
. Ined. Paspalum of eu о a. Hitchcock апа
227. Paspalum.
Contr.
Chase Library, — Departme nt, Smithsonian Insti-
tution, a on, D.
Cialdella, A. M., O. — & F О. Zuloaga. 1995. Re-
visión A las especies del género Paspalum (Poaceae
deris Paniceae), Grupo Bonplandiana. Darwin-
iana 33: 67-95
Clayton, W. i &
S. A. Renvoize. 1986. Genera Gramin-
um. Grasses of the World. Kew Bull.. Addit. Ser. 13:
1-38
la Silva, T. S., A. G. Burman & T. Sendulsky. 1979. Es-
pe ecies de Paspalum L.
da Ilha do Cardoso, estado de
Sao Paulo. Hoehnea 8: 11—28.
ا
Volume 89, Number 3
2002
Denham et al. 395
Paspalum subg. Ceresia
Davidse, & R. W. Pohl. 1972. Chromosome numbers
and notes on some M entral American grasses. Canad. J.
Bot. 50: 273-283.
— — . Chromosome numbers, meiotic
be havior and notes on — ‘al American grasses. Can-
ad. J. Bot. 52: 317—328.
&
. |“ 978. C ра — of tropi-
— — *
cal American grasses (Gramineae). . Missouri Bot.
Gard. 65: 637-64
Doell. J С. pu Graminae. /n: C. F. P. Martius к
Fl. Bras a 1-358. Munich, — Leipzig
Farris, J. 3 989. The retention index and hee rescaled
consistency — x. Cladistics 5: 417—419
‚ V. A. Alberte, M Kalle sjö, D. — & А.
G. РА 1996.
— jackknifing outperforms
ne ор Cladis es 12: 99-124.
Filgueiras, T. S. & ( Davide 1994. Paspalum biarista-
tum (Poaceae: — ае),
from Goiás, Brazil, and the second awned species in
Novon 4: 18—
&
«E
new serpe ntine е ndemic
E
the genus.
А О. c 1993. Ophiochloa. a
new endemic serpentine g Pani-
ceae) from the Brazilian cerrado vegetation. Novon 3:
310—317
erass genus (Poaceae:
Fliigeé, J. 1810 Graminum Monographiae. . .Pars |. Pas-
palum. Reimaria. V. Perthe es < J. Н. Besser, Hamburg.
Giussani, ү . J. H. Cota-Sánchez. F. O. Zuloaga « К.
. Ke — \ шейге ular phylogeny of the grass
А ub — ‘ade sae (Poaceae) shows multiple origins
of ( ооа Amer. J. Bot. 88: 1993-2012
A. 1993. Е anaing с — ‘ter — —
m n.
Re
^^
т
w
|
.C€ Mica :
07a. NONA version 2.0 for Windows.
,Onm-
— r pogran and doet 'ume ntatión — by the au-
thor, eee site: htt
97 H Pee - Wee version 3.0 ‘for > Windows. Com-
— r program and doc — eens by the au-
thor, website: idistic
Gould, F. W. & T. n. e 1967. КО num-
A of tropic ‘al American grasses. Amer. J. Bot. 54:
7 die »
Hac ke. - 1901. Neue Gräser. Oesterr. Bot. Z. 51: 233—
Manual of the Grasses of the Unit-
'er Publications, Ma w Yor
Holmgren, Р, K.. | . Holmgren & L. Eo tt. 1990.
Index He Pd Part | : The ا of the World.
New York Botanical Garden: m York.
Honfi, A. J. C. Quarín & J. Valls.
cariológicos en gramíneas
T ALS. Ii"
ed States, ed. eng
1991 [1990].
sudamericanas.
Darwiniana Ж 87-94.
` О. . Zuloags ga, Û. Morrone & А.
Ese obar.
canas (Poa eae Panic ol na) Darwiniana 35 —30.
Judziewicz, E. J. 1990. Family 187. Poaceae be А R.A
Gorts-van Rijn (editor), Flora of the Guianas. Serie
P ing sad Koeltz Scientific Publications. ues nig-
stein, Germany
Karis, P. О. 1995. Cladistics of the subtribe Ambrosiinae
(Asteraceae). = Bot. 20: 40-5
— T. J. 1990. The grasses e My cie Santa
+ via. fee Missouri Bot. Gard. 25-201
— D. 1991. The discovery and lon d of mul-
deg islands of most parsimonious trees. Syst. Zool. 40:
315
Mez. C. 1917.
pert. Spec.
VIII. Generis Paspali species novae.
Nov. Regni Veg. 15: 27-133
Re-
Morrone, O., ). Zuloaga & E. Carbonó. 1995. Revisión
del grupo zw emosa del | inue Paspalum (Poaceae:
Panicoideae: Paniceae). Ann. Missouri Bot. Gard. 82:
82
-116
‚ A. Vega & F. O. Zuloaga. 1996. Revisión de las
especies del género Paspalum L. (Poaceae: Panicoi-
deae: Candollea 51:
Paniceae). grupo Dissecta (s.str.).
103-138.
Jenham, S. S. Aliscioni & F. O. Zuloaga.
2000. Revtalón de ES especies de Paspalum (Panicoi-
deae: Paniceae), subgénero Anachyris. Candollea 55:
105-135.
Nash, G. V. 1912. P ae e © Tribe 5. Paniceae. North
American Flora 1 a 1-196
Nees ab Esenbeck, С. Еч 1829, Кога — ‘NSIS seu
) 1 Martius (editor) 2(1): 1-608
“М. a nter. 1993. On demie Cla-
distics 9: —
1805-1807. Synopsis Plantarum, 2 Vol.
Pilger, R F. 1929. Bemerkungen zur Systemalik der
Gathing Paspalum L. Repert. Spec. Nov. Regni Veg. 26:
: —
Pohl. |
. 1980. Gramineae. In: W. и г (editor), Flora
Costaricensis. Fie — Bot. 4: 350—392.
Renvoize, S. А. 1978. The genus — um Group Lorea
Segre — in the Gramineae: XLIII. Kew
Bull. 32: 419-428
1995. А new species — Mos ا (Gramineae)
from Bolivia Kew Bull. 50: ЗА
998. Gramineas de мена Сайна Print-
ing. н
Rodriguez Ж aL
nueva sección en el género Paspalum L.
1992. Pectinata Chase ex Rodríguez
(Gramineae
Ernstia 2: 21-2
1998. El — Ceresia (Pers.) Reichenb..
del género Paspaten L. (Gramineae) en Venezuela.
Ernstia 8: 7-50.
. B. R. Arrillaga de Maffei & P. Izaguirre
de Artucio. 1970. Gramíneas Uruguayas. Publicaciones
de : a Unive n de la Repüblic a, Montevideo.
Rua. H. У. 5. Aliscioni. In pre A morphology-
— dud a analysis of РЕ sect. Pectinata
(Poaceae). Syst. Bot.
К. 1948. Cariología ded gramíneas argentinas.
vista Кас. Agron. Veterin. 12: 51-67
Sendulsky, T. & A. G. m 197 8. Pasgalión species
of the Serra ds Ci Apó (1): A Bp to s study of
the Brazilian Poaceae. Revista Brasil. : 1-15
280. Paspalum species ni the Ser гга
А tan to bip study of the Brazilian
Bot. 3: 23-3:
йо к?
Saura. Re-
do Cipó (H):
Poaceae. Rev ista Brasil.
Smith, E. B... D. shades n wd Ң. M. Klein. 1982.
— as. In: R. iss (editor), Flora Hustrada Catar-
i Fas scículo Gram. 85. Paspalum até 115. Zea:
м
inense
910-1 10
оло. T. R. &
the genus Olyra at s J
olyra — ae: ies E Оте). Smithsonian
: 1-79
— 1989. A revision of
w segregate genus Parodi-
1962. A cytological study | E some Mexican
Torrey Bot. Club 89:
Tovar, O. 3. Las gramineas — Perú. Ruizia
13: 347-378
Trinius, €. B.
1826. De Graminibus Paniceis. Dissertatio
396
Annals of the
Missouri Botanical Garden
botanica altera. Impensis Ac a miae Imperialis Scien-
tiarum, Sa oli (St. Petersbu
— [8 18: 30. Species — iconibus et des-
c iptonibus illustravit. 3 vols. Impensis Academiae Im-
perialis Scientiarum, Petropoli.
1834, Panicearum genera retractavit specieb-
usque ае illustravit Mé е Acad. St.-Pétersb.
Sér. € . Phys 3(2): 89-35:
— 0. & 0. ы. In prep. Ңеуїзїбп de las
especies de Paspalum para América del Sur Austral
(Argentina, Bolivia, sur de Brasil, Chile, Paraguay у
Uruguay).
APPENDIX |
List of accepted species of Paspalum subg. Ceresia.
1. da en aspidiotes Trin.
2. P. biaristatum Filg. € Davidse
3. buchtienii Hac
4. P. burmanii Filg., Morrone & Zuloag
5. P. cachimboense Davidse, Morrone & EL
6. P. carinatum Humb. & Бор}, ex Flugg
7. P. ceresia (Kuntze) Chas
8. Pe
9. P. cymbiforme E. Fourn.
10. P. ees ex Trin.
П. P Morrone & Zuloaga
12. P. hete
13. P. biis ciis Flüggé
. P. imbricatum Filg.
15. P. lanciflorum Trin
lo. P. D un Davidi & Filg.
17. P. malmeanum Ekman
18. P. ае Fi
19. P.
20. Р
21. P. phyllorhachis Hack.
22. P polyphyllum Nees ex Trin.
. P. reticulinerve Renvoize
24. P. stellatum Humb. & Bonpl. ex Flügg
25. P. trachycoleon Steud.
eucomum
goyase nse Dav —
trichon
APPENDIX 2
Index to exsiccatae. Each :
listed by the first collector. The number in parentheses
refers to its associated species (see Appendix 1). An as-
terisk (*) indicates a voucher for the o study.
; Acosta Solís 21454 (13); Ahumada
3541 (24); Alba 53-23 (19), 53-96 (6 M Allem 1558 (6),
2007 (6), 2236 (24); Almeida 572 (19); Alston 7569 (19);
Amat 3 (15); Anderson 1120 (24), 8286 (24), 9413 (22),
9421 (24), 9969 (6), 10304 (22), 35213 (19), 35798 (22):
Anisits 2047 (24); — 104. (19), 144 (22), 345 (24);
Archer 629 (13), 3843 (24), 4995 (22); Arechavaleta s.n.
(22); Argent 6460 (22), a (19); Aristeguieta 238 (25),
4219 (6); Arsène 2813 Asplund 6531 (13), 7593
* 10498 (13), 10633 no 12014 (13); Aynard 6019
ail 100 (25); Bailey 112 (2: h — (25); ۰
(24), 90 (22), 90a (22), 126 (24), 127 (25): Balick 1
(25): i 6923 (13); Bang 265 (: E B: on sa aa
3032 (19); Barclay 5703 (13); Barlow 30/1173 (19); Bar-
reto 46 (24), 529 (22), 1483 (22). 1577 (24), 2952 (19):
Basualdo 2162 (24); Beck 4025 (7). 8714 (3), 12330 (24).
19948 (3), 20552 (19), 20564 (6); Bentacur 1510 (19);
Berro 2680 (22); Bertoni 5406 (24); Black 47-1777 (6).
51-12029 (19), 51-12148 (22), 51-12291 (6); Blake
specimen is alphabetically
TS 7
T
—
22
7445 (19); Blydenstein 734 (19), 736 (24), 847 (6), 856
(19), 900 (19), 939 (8), 1037 (19), 1082 (6), 1188 (24),
1468 (15), 1562 (19); Boeleke 16646 (13); Borges 189
(22); Brade 19638 (8), 20377 (22); Breedlove 3313 (13),
13822 (9), жүр (9), 39106 (9), 42481 (25), 53954 (24),
54302 (12), 54468 (25), 54564 (25), 54586 — 54596
(12), s.n. (2: 5), Brooks 70 (24), 125 (7),
reck 2 (6); Buchtien 447 (3), 3629 (3), < an. (i "
B 3 V0495 (24), V0530 9— Bure hell 2 (22).
4574 (22), 4081 (22), 5769 (22), 7061—2 (25), 7124 (25),
7295 (25); Burkart 5335 (13). 10211 (13), 11014 (13),
16984 (13), 18242 (24), 21610 (24); Burman 393 (24).
Caballero Pardo 77 (7); Cabrera 2278 (8), 22730 (7),
28254 (24); Calderón 2272 (24), 2708 (8). 2760 (1); Cal-
lejas 6886 (15); Calzada 17051 (19); € ramp E-2743 (13),
2332 (7); Capell s.n. (22); Cárdenas 515 3s 3410 (13),
3606 (22); Carnevali 1083 (24), : 3200 (22),
=
*—
=
(24), 9148 (22). 9212 (19), 9250 Ts 9953. (24), 9205
(24), 9308 (24), э (21), 9352 (22), 9367 (22), 9369
(22), 9559 (22), 9917 (22). 10230 (19), 10249 (19), 10276
(19), 10306 (6), 10850 (19), 10595 (6), 10655 (19), 10740
(22), 10872 (24), 10907 (24), 11200 (10), 11543 (24),
11549 (22), 11602 en 11768 (24), 11962 d 12014.
oe 12028 (24), 12085 (2 5), 12088 pies y
2292 (25), 12294 (13); Chávez 3217
үз Claussen s.n. (24); Clayton 1631 (8), rm (6); Clewell
3514 (2: 4: C onzali 4364 (24); Coradin 696 (6); Costa Sac-
co 133 — 21469 (25); Cuatrecasas 3001 (13),
25702 (13).
p D'Arey 141 (22), 2271 (22), 2659 (25), 4082 (6), 13527
(15); Daly 2171 (22); Daniel A-2 (19), 60 (6); Davidse
3065 (25), 3082 (12), : 4). . А
4514 (25), 4615a (6), 4708 (1), 5109 (19), 5242 (19),
5321 (19), 5597 (13), 10126 (24), 10655 (22), 10967 (22),
10985 (22), 11028 (22), 11123 (22), 11148 (22),
(22). 11316 (22), 11380 (22), 11401 (22), 11407 (22),
12087 (6), 12124 (6), 12158 (25), 12159 (24), 12276 (24),
. 15652 (19), 16117 (19), 16775 (6),
55 (6), 19314 (15). 19817 ө) 19487 (25),
(6), 20544 (19), 21162 (25), 21312 (25), 21370 (13),
22047 (15), 22944 (1), 28159 TEN po 37 o. 35448 (13);
Dawson 14142 (24), 14215 (24), 14863 (2:
192 (22); de Jesus 3 (10), 16 (24); de Lima. s.n. (22); de
Paula 65 (24); Dedecca 302 (22); Dias 67 (10); Dinelli
12876 (7); Dombrowski 780 (19), 4570 (22), 5984 (22),
6488 (19), 6715 (19), 9037 (19), 9784 (19); Drew E-747
(24); Dusén 26a (22), 283 (22), 2770 (19), 2844 (19),
2844a (19), 3657 (22), 3659 (22), 4023 (24), 8011 (24),
9109 (8), 10412 (6), 10611 (19), 10613 (19), 13256 (19),
1: 3805a (22), 14489 (22), 15727 (19), 16057* (19), 16174
2), 17407 (19), 17934 (6), 18006 (24),
(10), s.n. (22); Dutra 464 (22).
Echeverry 1251 (13), 2312 (15); ene 639 (9); Egler
1262 (15): Eiten 3777 (7), 4417 (24 8 (7), 6746 (19),
6760 (19), 7626 (22), 9151 (19): а 594 (24), 596
2274 (12), 2302 (24), 2401 (12), 2484 (24), 13404
(12). 14018 (12); Ellenberger 1369 (22); Enamorado 23
(9); Escalona V274 (24); E (22).
Fassett 25502 (13); Fe Тено 6462 (22); Fendler 1697
(25), 1698 (12), 2533 ( st Fernández 89 (24); Fernández
Casas 6335 (24); Ferre 3 (13): Fiebrig 664 (24)
2061 (7), 3082 (13), 497 О (24), 4978 (24), 6222 (22);
Filgueiras 2131 (14), 2184 (1: 4), 2280 (7), 2342 (16), 2408
(2), 2409 (2), 2455 (19), 2477 (4), 2780 (16), 3260 (22),
3392 (16), 3317 (14), 3567* (25); Filipovich 398 (13);
Volume 89, Number 3
2002
Denham et al. 397
Paspalum subg. Ceresia
Flores Flanco 3643 : 3); — 224 (2), p — 600
24 Ape (1); Freytag s.n.
6 (24 ); Fune ke 7 14
García Barriga 4993 (19), Dos п» ji
‚ 4390 (25); Gentry 15081 (19), 7
G-101 (15); Giraldo-Cañas 2560 (6),
2260 (22), 4309 (22), 6775 (22), 7980 (22), 13334 (24),
16548 (22). 17406 (10), 17410 (22). 17411 (22), 17414
(24). 17940 (22). 18607 (22), 18644 (22), 18674 (22),
20070 (21), 20076 e 20078 (21). 20079 (25), 20082
(22). 20083 (22), 2 4 (15), 22427 (19). 22437 (22
22441 (11). 22443 a n 22544 (15), 22544b (14), 22545
. 22547 (6), 22549 (24), 22551 (10),
22553 p.p. (24), 22555 (10). 22556 (6). 22551 (24).
22558 (22), 22559 (22), 0 — 22561 (22), 22562
(22), 22576 (12), 22577 (25), 22949 (24), s.n. (24); Goés
s.n. (22); Goodland 923 (15): — rger 1043 (10); Gra-
ham 520 (19); Grignon 84257 (13); Guala 1320 (24), 1450
(22); Guánc A > 3018 (1); Gutiérrez 655 (17).
Haase ) (19); Hahn 5592b (15); Handro 97 (22):
Harle "y T ^ (19). 10480 (19), SH209 (22); Hassler 1039
. 1058 (24), 9259 (24), 9261 (22), 926la (22). 9380
. 10209 (22). 11058 (24), 11501 (22), 11561 (22).
11916 (22); Hatschbach 5461 (22), 6864 (22). 7151 (22).
8965 (22). 11110 (22), 12097 x 2h 16134 (22). 25048 (19).
25264 (19), 32479 (19), 33152 (19), 44562 (22). 15097
(19), 47650 (24), 48811 (19), 64890 (22), 66690 (22):
Haught 2709 (19); Hazlett 994 (24); Heithanus 163 (19)
Heringer 4615 (15), 7249 (22), 7382 (14). 10432 (22):
Hermann 10995 (19), 11089 ^а Hernández 2283 (19);
He Du Xolocotzi X-623 (12); Herter 1755 (22): Heyde
6211 (1 Hicken s.n. (24); Hill 1548 (19); Hinton 1440
(13). mad ө, 11400 (13); Hitchcock 6355 (13). 0837
7297 (13). 8186 (13), 8297 (12). 8946
Gardner 4030
Gifford
—
(24); Holway 98 (22),
Hostmann 1306 (6): dun 9643 (1 b
(24). 6012 (22). 6246 (22); Hutchinson es (13).
(7).
1); Hunt 5550
5194
(24); Idrobo 11636 (13); Hus 342 (13):
inen 7 739 (6), 5163 (24). 5171 (22).
, 6205 (22), 6579 (19), 6758 (19).
7219 (22), 8001 (24). 9495 (22).
Ibarrola 2834
Imaguire 605 (22);
56
6800 (19). 7069 (22 2),
10737 (11). LUD. 13241 a D. р 15328
(10). 15347 ee 15928 (24), 15929 (24),
16232a (22), 16465 (24), 17256 (22), 17314 (24), 17617
(24), 17854 (22), 17967 (22), 19486 (22). 27667 (О),
28353 (22); Isabell s.n. (7).
Jangoux 10153 (19); Joergensen 1765 (13). 2882 (24).
4568 is 1); Joly 1456 (24): Jorgensen 52 (24): Jürgens G-
299 :
E 087 (24), 904 (24), 1795 (22), 1952 (24). 1996
(24). 201 1B (24), 2024* (17), 2076 (17). 2458 (24), —
). 2474* (24), 2477 (24), 247 (24). 278
(19), 2795 (22), 2823 (6), 4152 (7). -
6056 (2: 1). 6541 (23); Killip 1555 (12).
‚ 19770 (13), 34321 (19); Klein 3916 (19): Koninck
157 (25 ›); Kral 69245 (19), 72050 (1), 75939 (11); Kra-
povickas 14972 (24), 25202 (24): Kuhlmann 1667 (17),
1668 (17). 1671 (23). 1674 (15). s.n. (22); Kummrow 2204
(22).
Lazor 3350 (19); Leavenworth 1649 (1 2: : eehmann 979
19); Leonard 7526 (24). 7537a (24). 3 (12), 895:
„ 8957 (12); Liebmann 225 (9); p 19350 (15):
Ama 58-3056 (24); Linares 3663 (9), 3904 (9); Linden
=
= x
ы
5); Llano 47 (6); Llatas Quiroz 2322 (13); Loefgren
gos 351 Pg" 1121 (12), 1543 (8): hado de
Campos 111 Machuca Nuñez 4040 (1:
101 (22), 908 d 19226 (24); Maguire — (15); Mal-
me 965 (22). 1562b (12), 2226 (22). s.n. (22):
1254 (13); Marin 851 (13); Martius 267 (22): Martínez F-
1953B (13); MeVaugh 19098 (13); Mendoga 2293 (22):
— 2943 (24); — = (24); Меуег 2090 (22),
; Miranda 3883 : Molina 8702 (25). 12041
“tz? (24): Montes — ). 1953 (24). 9)
| 1920 ( 24), 15353 (24), 16250 (24), n (24). 27042
(24); Montoro s.n. (22); Moraes 1713 ; Moran 5505
: Mori 13303 (22), 17298 (6); ا T (25); Mor-
rone 205 (24); Mostacedo 314 1783 (23), 1862 (1):
Müller 2036 (13); Mutis 5453 (19), 5537 (13); Myers 3338
E
(19), 3404 (24).
Nee 8166 (24), 26848 (3), 40372 (7), 46904 (19).
19412 (22); Nelson 3424 (24); Nicora 4700 (22), 4708
(24). 6352 (24); Niederlein 570 (22); Norrmann 159 (24):
Nunez 7175 (13).
Oldenburger 5 (15); Oliveira 305 (7). 598 (7). O15 (2).
623 (2). 630 (2), 638 (16), 700 (16). 719 (12), 758* (2).
763 (19): Ollgaard 90721 (7); Orozco 182 (19): Orth 1059
22). 4664 (24), 7070 (24): Partch 09- 91
3): Pedersen 1032 (22), 2611 (24). 3783 (24).
(22), 13371 (22), 15713 (24); q nnell 8174 (19);
net 496 (22); Peterson 9478 (7), 11887 (22). 1 1801 (2 t);
Pinheiro 142 (7), 517 (7), € A (7); Pinto d (19), 894
(19). 1066 (19). 1141 (19), 1467 (19). 1550 (19); Pire
14649 (24); Pires 2290 (24). 16684 (24): — r 214 (25).
1788 (25) 4351 (19), 5020 (24), 5902 (25). 0033 (25),
7487 (24). 24 ‚ 11326 (24),
11998 (2 D
19132 (15); Pohl 714 (19), . 2833 (24).
(24). 12274 (6), 12361 (19), 13531 (9). 13761 (25). 13850
(24). 13922 (19); Porter 4300 (19); Prance 13823 (1): Pur-
pus 2002 (13); Purseglove 6514 (6); Pursell 8580 (25).
Quarín 2235 (24), 3340 (24), 3716 (24). 3852 (24).
1043 (21); Quintana 165 (24); Quiroga 38 (24).
Rambo 40511 (22), 40690 (22), 40804. (22), 53900
(22), 54017 (22), 54991 (22); Ramia 1009 (24); Ramirez
961 (15). 1789 (24): Ramos 182a (6); Ratter 520 (22).
820 (21). 1915 (24). 1969 (19), 2883 (24), 2884. (25).
3335 (24); Reeder 3313 (13); Reeves 194 (25); Regnell
111-1345 (25). HI-1344 (24), 111-1335 (19), 1-466 (22); Re-
ineck 431 (22); Reitz 10652 (22), 10659 о 10081 (22),
10906 (22), 10933 (22), 12477 (22);
1038 (13). ibi — 5409 (19); Riedel 966 (12),
(22), 1048 (1), 2143 (22), s.n. (1), s.n. (12), s.n.
(24); — ra L-L41 (19), 3 (19); Rizzo 4285 (22);
Robinson 3189 (22); Rodrigues 662 (6); Rodriguez 206
(24); Rojas 2757 Qu. 9569 (22), 13404 (22), 13933 (22):
p Rombouts 308 (19); Rondon 2135
24); Rosengurtt B-3197 (22), B-5749
22), B-5883 (24), B-7397 (22); Busia 1200
(25): Rzedowski 17636 (13).
2332 (7), 2341 (13); — M (24):
(22) 4215 (22), 5 (22);
t (13); Schenck 3155 m M
1323 (24). 1403 (24). 1436 (24), 6207 (24); Schmidt 53
(22), 136 (13); Schreiter 9251 (13), 9256 WA ч 300
(24). 7220 (24). 17275 (22); Schwacke 8444 (2
2214 (22), 5539 (24). 5703 (24). 10003 (24), тл (24);
Sellow 1232 (19), 1239 (19), 1240 (22), 4416 (22). 5686
(24). s.n. (19), s.n. (22). s.n. (24): Semir 380 (19); Semple
Romanezuk ' 12
(9) Rusby 1014
Sánchez Vega
Saravia 14006 (13),
Schaffner 120 (13), 15
Annals of the
Missouri Botanical Garden
150 (19); Sendulsky 1239 (10); Sermeño s.n. (13); Shaw
s.n. (19), s.n. (6); Silva 114 (24); Smith 142 (24), 143 (12).
2170 (25), 5907 (22), 6316 (22), 6912 (24), 8445 (19).
11529 (24), 12013 (22), 12106 (22), 14462 (6). 14579 (8).
14858 (8), 15569 (19), 16123 (22); Soderstrom 1124 (6);
Sohns 427 (13); Soria 4442 (24); Soriano 986 (13);
—
~
; Souza
5548 (7); Spegazzini s.n. (22); Standley 13185 (12), 56275
4); Steinbach 3015 (3), 3712 (22), 5162 (22), 5385
6618 (13), 6808 (22), 7103 (24); Stevens 24761 (13), 25701
(13); Steyermark 29550 (9), 32312 (24), 42586 (25), 55
(25), 59150 (19), 75259 (19), 121984 (25), 127710 (24).
127714 (25); Stockwell 893 (13); Stuckert 14159 (13),
18694. (22); Swalle n 3074. (19), 3643 (6), 3687 (6). 3772
4078 (6), 4762 (6), 8344 (19), 8361 (22),
3418 30 EA 8518 (22), 8584 (22), 8613 (22).
8639 (22), 8698 (22), 8772 (22), 8784 (22),
8975 1/2 (22), 8994 (22), 9033 (22), 9083 (22), 9087 —
9132 (22), 9150 (22), 9159 (22), 9181 (22), 9540 (2
9590 (24), 10747 (9), 10809 (9). 10810 (12), 10993 (25
R ‘ ).
; Tamayo 1084 (25), 1867 (22), 2046*
(25), 2281 (13), 2881 (2 4), : 3198 (15), s.n. (15); Tate 157
* (12); Triana 761 (25);
(24
Ule 7919 (24), 80 32 (24).
Valeur 498 (24); от 8340 (6); van der Werff 7863 (22):
Vargas 2459 (7), 19047 (7); Venturi 8320 (7), 9905 (7);
Viei v 54 (22); ee Carenzo 1818 (13); Vincelli 1146 (19).
Warming s.n. (25); Webster 29732 (19); Weddell 1699*
(8), 2556 (10); Widgre n s.n. (22), s.n. (19); Williams 10533
(24), 13088 (15), 14704 (24), 17344 (24), 42957 (25); Wood
4273 (22), 4306 (25), 7852 (13); Wurdack 27 (15).
Zardini 26848 (24), 51117 (19); Zanoni 31952 (24); Zu-
loaga 888 (22), 910 (24), 2383 (22), 3835 (11), 3839 (19),
4379 (6), 4413 (15), 4587 (22), 4606 (22), 4654 (22), 4697
(22), 4874 (22), 4904 (22), 4930 (22), 5371 (24), 5671 (24).
5788 (24), 5791 (22), 5885 (13). 6585 (22), 6817 (24).
6944. (22)
T
APPENDIX 3
List of specimens for outgroup taxa in the cladistic anal-
ysi є.
Anthae "nantiopsis | fie brigii L. Parodi
ARGENTINA. Jujuy: Dpto. Dr. M. Belgrano, Zapla, Mina
9 de Octubre, Cabrera et al. 32666 (SI).
Panicum
I — laxum Sw.
ARGEN PIN A. Corrientes: Monte Caseros, monte del ar-
royo El Seibo, Nicora 4516 (SI). Misiones: Dep. Iguazú.
Cataratas tar Cabrera et al 28942 (SI). Santa Fe: Río
Paraná, Canal Viejo, Job 710 (SI).
— ит — Kunth
J.S.A. Arizona: Box Canyon Road, Gould 2889 (SI).
Paspalum € NS Dissecta
Paspalum acuminatum Raddi
ARGENTINA. Tucumán: Dpto. ida Loc. Monteros,
Lillo 38377 (SI); Dpto. Leales, Loc. Leales, Venturi 699 (SI).
Entre Ríos: Dpto. Feliciano, río Guayquiraró, Ruta 12,
Burkart et al. 23250 (SI).
Paspalum — tum (L.) 1
: N side a Gulf Shores, Kral 51314 (MO).
Louisiana: "Cher eniere Lake, Burkunas 605 (MO).
P uns Group Eriantha
— erianthum Nees ex ‘Trin.
LIVIA. Santa Cruz: Ñuflo de Chávez, Est. Las Madres.
Killeen 1369 (SI). BRAZIL. Distrito Federal: Area da
Proflora, Filgueiras & Zuloaga 2207 (SI); Brasília, jen
Sul, Zuloaga 3836 (SI)
Paspalum guttatum Trin.
BRAZIL. Distrito Federal: — en Brasília & Sobradin-
ho, Irwin et P 9163 (MO). Goiás: s.l., Glaziou 22442 (G).
Paspalum — ntum Trin.
BRAZIL. ‚ Go of Cristalina, /rwin et al. 9858 (MO).
i Santuario, Zuloaga &
=
as Gerais: — da Piedade,
Morrone 4676 (SI).
Paspalum Group Notata
: Tavarory, 3 km S from administra-
E of río Paraguay, Zardini & Tillería 29180 (S1):
lta — — 6277 (SI). Pte. Hayes: Estero Pa-
tiño, ruta Trans Chaco, km 164, Quarín 4049 (SI).
— нати Trin. ex Doll
ARGENTINA. Corrientes: "el del Timbof, о de
Monte Caseros, Nicora 4629 (SI). BRAZIL. Río Grande
o Sul: Centro Agronómico de Guaiba, Burkart pes (Sh.
URUGUAY. Durazno: La Paloma, Schroeder s.n. (SI).
jos notatum Flüggé
ARGENT pa Córdoba: Santa María, Alta Grac la, E
zker 1178 orrientes: San — 19 km S de (
alí, Ruta 9 Maas 1837 (SI); Monte Caseros, ai
el Urug , Nicora 4587 (SI). Misiones: Dpto. Apóstoles,
aida Burkart — (SI).
Paspalum Group Rac
— — How. & Bonpl. ex Flüggé) Kunth
NTINA. Jujuy: Dpto. Dr. Manuel Belgrano, faldeo
del rh Se nde то pos de Lozano, Deginani & Cialdella
135 (SI); Dpto. Capital, Lagunas de Yala, Cabrera et al.
306706 (SI); Termas de Reyes, Burkart et al. 11218 (SI).
Paspalum — Scribn. & Merr.
MEXICO. Hidalgo: near — Station, Pringle 8891 (SI).
VENEZU A. Lara: Distrito Jiménez, 4—5 km of Sanare,
Davidse & González 21366 (SI).
Paspalum racemosum Lam.
PERU. Lima: Chancay, Lomas of — S of Chancay,
Stork & Vargas 9344 (SI). Trujillo: La Libertad, Cerro Cab-
ezón, Sagástegui & Mostacero ze (SI); La Libertad, Bar-
raza, Mostacero & Ramírez 690 (S1).
=
APPENDIX 4
Index to scientific names. Synonyms appear in italics.
Biaristata, sect. 338, 352
Jeresia 337, 352
Ceresia membranacea 363
Ceresia, sect. 338, 339, 352
337, 338, 339, 340, 344,
346, 348, 349, 350,
Ceresia, subg.
351, 352
Ceresia, subsect. 352
Panicum ceresia z:
P. obtectun
Paspalum aspidiotes
|
у. 339, 340,
352, 354, 381
* 351,
P. aureolatum
P. biaristatum
377
338, 339, q 346, 348,
35 i
356, 357, 379
P. bicilium 386, 38€
P. blepharophorum 337, xs 373, 388
P. blepharophorum var. 386
tenue
P. buchtienii
338, чү 340, 344,
347, 351, 352, 357, 375
Volume 89, Number 3
2002
Denham et al.
Paspalum subg. Ceresia
399
P. burmanii
P. cachimboense
P. carinatum
P. ceresia
P. ceresioides
P. ciliatum
P. contractum
P. cordatum
P. cujabense
P. cymbiforme
P. distichophyllum
P. echinotrichum
P. elegans
P. erectifolium
P. eucomum
P. eucomum var. pilosior
P. govasense
P. gracile
P. heterotrichon
P. heterotrichon var. pau-
cispicatum
P. humboldtianum
P. humboldtianum var. ele-
rantissima
P. humboldtianum var.
stuckertii
P. imbricatum
P. kappleri
P. lanciflorum
339, 346, 348, 351. 352.
360, 380
359, 360, 38(
339, 340, 344, 346, 348
352, 3 » 3 L 37i
362, : 70
338, 340, че 348, 351,
352, 363. 391
7l
НЕ
338, 351. 377
338, er 346, Mp m
354, 365. 378, :
389
338, 339, 1T 346, 347,
351. 352
373, 388
377
352. 363
337. 338, 346, мы чш
352. 307, 366
т 9]
m. 2 * 351. 352.
3 K
338, 339, 346, 351, 352
371, 383. 392
311
338, aye eo 344, 340.
347,
360, ш = 375,
376, 380, 388
375
338, 375
339, е awe 340, 348,
351
361
338, 339, ЗЛО. 346. 352.
301. 377
P. longiaristatum
&
P. macroblepharum
I
P. malmeanum
P. membranaceum
P. membranaceum var.
aequiglume
P. membranaceum var.
inaequiglume
P. niquelandiae
P. pectinatum
P. petrense
P. phyllorhachis
P. piligerum
P. polyphyllum
P. reticulinerve
P. setiglume
P. soboliferum
ns
P. splendens var. sphacela-
tum
P. stellatum
P. stellatum f. паша
monostachyum
P. stuckertii
P. trachycoleon
P. wagenerianum
Pectinata, sect.
Tricholaena obtecta
378, : 19
386, 384
3: " 3 7 351,
379, 391
352. 362.
90
389
389
338, 339, 340, ж
348, 351, 35:
303. 389, 391
ч 33. 339, 346,
352, —
386. :
301,
389
338, 339, 340,
354.
37 |
377,
340.
18. 351.
383,
Xx
352. 369,
331. E. 363
363
363
339. 346, 348, 351. 352.
2 79. 380
337, 338, 340, 346. 351.
352. 365, 378, 380
339. 348, 351. 352. 371.
383, <
3: ES gt 346, 348. 351,
352, 371. 383. 386, 392
377
337. 338, 339, 346, 317,
4 Е = 367. 375
"5 gr ae 346, 351.
352, 354. 388. 389
338. 354
338, 375. 376
361. 362
338. 367. 370
389
337. 338. 340, 346, 351.
350. 352
380,
39
PHYLOGENY OF THE TRIBE Alan W. Meerow,? Charles L. Guy,’
HYMENOCALLIDEAE Qin-Bao Li,* and Jason R. Clayton’
(AMARYLLIDACEAE) BASED
ON MORPHOLOGY AND
MOLECULAR CHARACTERS!
ABSTRACT
The gene ric limits of Hymenocallis have been variously proposed by different taxonomic workers, often without
discussion or data. The genera Leptochiton, Ismene, Elisena, and Pseudostenomesson have been included with Hymen-
ocallis, lumped together as the genus /smene, or maintained as distinct genera. Recent cladistic analysis of plastid and
nrDNA for Aar рари — a distinct tribe Hymenocallideae. Cladistic analyses of morphology, and plastid
(trnL-F region) and nuclear sl DNA (ITS) are prese е alone and in combination for the tribe. Leptochiton is
sister to the rest of the genera in the tribe in all analyses. While Hymenocallis is always resolved as monophyletic,
Ismene is variably paraphyle lic or monophyletic. The enda sequence data produce the most resolved and best-
— jd phyloge ny, wherein Hymenocallis and Ismene are monophyletic sister genera. These data support an origin
for the tribe in the Andes, with vicariant distribution of the largely Mesoamerican Hyme — Formal recognition of
Ismene subg. Elisena and Pseudostenomesson is established.
Key words: Amaryllidaceae, cladistics, molecular systematics, phylogeny.
Systematics of the genus Hymenocallis Salisb. — menocallis, subg. Ismene (Salisb.) Baker ex Traub
(Amaryllidaceae) and its allies have defied precise (including Leptochiton), subg. Elisena (Herb.)
systematic understanding at both the specific and Traub, and subg. Pseudostenomesson (Velarde)
generic levels (Flory, 1976: Meerow € Dehgan, Traub. Traub (1980) later reduced these subgenera
1985). The genera Hymenocallis and Ismene Salisb. to the rank of section without explanation. Ravenna
were established by Salisbury (1812) for the Neo- — (1980) in his description of H. heliantha (= Lep-
tropical species with fleshy seeds originally as- tochiton heliantha (Ravenna) Gereau € Meerow)
signed to the Old World genus Pancratium L. The suggested that subgenera /smene (including Lepi-
zygomorphic-flowered Elisena was described by dochiton). Elisena, and Pseudostenomesson should
Herbert (1837), who recognized Hymenocallis and probably be all recognized as the genus /smene,
Ismene as distinct genera. Baker (1888) subsumed distinct from Hymenocallis. Meerow and Dehgan
Ismene within Hymenocallis but retained Elisena as (1985) suggested that Pseudostenomesson might
distinct, as did Pax (1890). While Stapf (1933) warrant recognition at the rank of genus due to its
treated A. quitoensis Herb. as a species of Pamian- extreme phenetic divergence. (funnelform-tubular
the Stapf, Sealy (1937) considered the species to perianth) versus the “pancratioid” flower of Lepto-
exhibit sufficient’ morphological divergence to be — chiton, Ismene subg. Ismene, and Hymenocallis.
recognized as a monotypic genus, Leptochiton Sealy.
€
©
—
’ancratoid” floral morphology refers to a large.
Hutchinson (1934, 1959) retained both Elisena and white, fragrant, crateriform flower with a conspic-
Ismene (presumably including Leptochiton) as dis- uous staminal cup (ef. Pancratium L.). This type of
tinct. Velarde (1949) established the Peruvian ge- flower appears to be adapted for sphingid moth pol-
nus Pseudostenomesson for a fleshy-seeded species — lination (Bauml, 1979; Grant, 1983; Morton, 1965).
originally described as Stenomesson morissonii Var- Meerow (1990) treated Leptochiton as a distinct ge-
gas as well as one new species. Traub (1962) rec- nus and recognized Hymenocallis and Ismene (in-
ognized all four erstwhile genera as subgenera of cluding Elisena and Pseudostenomesson) as distinct,
Hymenocallis in his synoptic treatment: subg. Hy- a treatment followed by Gereau et al. (1993) and
е work was supported in part by NSF Grant DEB 9628787 to gi M and (
SDA-ARS-SHRS, 13601 Old C de < Miami, Florida 33158, U.S.A., and — hild ‘Tropical Garden, 11931
ом г tle — Miami, Florida 3315
Iniversity of Florida-IFAS, De iube nt " Environmental Horticulture, 1545 Fifield Hall. Gainesville, Florida
—
USDA. ARS- SHRS, 13001 Old Cutler Rd., Miami, Florida 33158. U.S.A.
ANN. Missouri Bor. GARD. 89: 400-413. 2002.
Volume 89, Number 3
2002
Meerow et al.
Phylogeny of Tribe Hymenocallideae
Meerow and Snijman (1998). No cladistic analysis
has focused exclusively on testing the validity of
this treatment, although at least one representative
of each subgenus was included in overall molecular
studies of Amaryllidaceae (Meerow et al., 1999,
2000a).
Hymenocallis and its allied segregate genera are
entirely Neotropical in distribution [a single West
African taxon, Н. senegambica, was treated by Sea-
ly (1954) as an early adventive introduction of H.
caribaea]. Hymenocallis sensu stricto, with 50 to 60
species, is chiefly Mesoamerican and extends into
the West Indies and the southeastern United States.
It is sparingly represented in northern South Amer-
ica. Leptochiton Sealy (2 spp.). Ismene (ca. 7 spp.).
Elisena Herb.
Velarde (2 spp.) are all endemic to the Central An-
—
2 to 4 spp.), and Pseudostenomesson
dean region of South America. Hymenocallis, Is-
mene, and Leptochiton are contrasted in Table 1
Hymenocallis and allies have usually been allied
with Eucharis Planch. in the tribe Eucharideae
(Hutchinson, 1934, 1959; Traub, 1963; Dahlgren
et al. 1985; Miiller-Doblies € Miiller-Doblies.
1996). Meerow (1989, 1995) argued that the link-
age of these genera, largely through the perception
that both lineages shared a fleshy seed, was mis-
construed, and proposed that either subtribal or
tribal recognition of Hymenocallis and allies was
Miiller-Doblies Miiller-Doblies
(1996) placed them in Eucharideae subtribe Hy-
warranted. and
menocallidinae, while Meerow and Snijman (1998)
recognized a distinct tribe, Hymenocallideae. Fam-
ily-wide analysis of plastid sequences (Meerow et
al., 1999) and nrDNA analyses of the monophyletic
2000a)
support a distinct Hymenocallideae as sister to the
American clade of the family (Meerow et al..
newly recognized tribe Clinantheae (a segregate of
the former Stenomesseae), but complete resolution
of the intratribal relationships is not apparent in
these large analyses. Both tribes are subclades of
a well-supported, Andean, tetraploid clade of gen-
era.
In this paper, we present phylogenetic analyses
of morphological and molecular data for the tribe
Hymenocallideae, and seek to clarify the relation-
ships within the tribe.
MATERIALS AND METHODS
SAMPLING
Sequences for the plastid trnL-F region were
newly obtained for H. eucharidifolia, which, along
with H. latifolia, was used as an exemplar taxon of
Hymenocallis (Table 2). Previously cited sequences
were used for one species each of the three sub-
genera of Ismene, one species of Leptochiton, and
the outgroup Pamianthe peruviana (Table 2, Mee-
999). For ITS and the morphological
data matrix, we increased our sampling with an ad-
row et al.,
ditional four species of Hymenocallis and two ad-
ditional species of Ismene subg. Ismene (Table 2).
The aligned sequence matrices are available from
the first author (miaamOars-grin.gov).
MORPHOLOGICAL DATA
Morphological and cytological character. state
data were derived from the following sources: Traub
(1962, 1980), Sealy (1954), Flory (1976), Velarde
(1949), Bauml (1979), Meerow and Dehgan (1985);
from examination of living material in research col-
lections at the USDA, Miami,
vations of Hymenocallis, Ismene, and Leptochiton
Florida; field obser-
species; and examination of herbarium material.
The morphological matrix consists of 12 species
representing 4 genera and 23 characters (Tables 3,
t)
SEQUENCE DATA
The trnL-F (trnL intron and spacer region be-
tween trn and trnF) matrix consisted of 6 taxa and
906 base positions. The nrDNA ITS sequence ma-
trix (IISI. 5.85 intron, ITS2) consisted of 12 taxa
and 636 bp.
DNA EXTRACTION, AMPLIFICATION,
PROTOCOLS
AND
SEQUENCING
silica gel
Genomic DNA was extracted. from
dried leaf tissue as described by Meerow et al.
(2000a). The trnL-trnF region was amplified using
the primers of Taberlet et al. (1991) as described
by Meerow et al. (1999). Amplification of the ri-
bosomal DNA ITSI/5.8S/ITS2 region was accom-
plished using flanking primers (185, 265) ABIOI
and AB102 (Douzery et al., 1999), and the original
White et al. (1990) internal primers ITS2 and 3 to
amplify the spacers along with the intervening 5.95
sequence, as described by Meerow et al. (2000a).
Amplified products were purified using QIAquick
(Qiagen, Valencia, California) columns, following
manufacturer's protocols. All polymerase chain re-
actions (PCR) were performed on an ABI 9700 (Ap-
plied Biosystems, Foster City, California).
Cycle sequencing reactions were performed di-
rectly on purified PCR products on the ABI 9700,
using standard dideoxy cycle protocols for sequenc-
ing with dye terminators on either an ABI 377 or
ABI 310 automated. sequencer (according to the
manufacturer's protocols; Applied Biosystems, Fos-
ter City, California).
402
Annals of the
Missouri Botanical Garden
‘Ob ‘OB-€EZ = uc
3[qegrre^ mq
пошшо) jsou
Ob ‘OF = UZ juasqy
9p — uc
juasqy
Ju9sq y
OLIPPOL
Ju9sq y
juasa1g
01-с
тс
0791
1y3regs pue 3u0| juour
-e[y әәлу шим pedis jou *odeus
ш әүдеивл pue jews 10 93181 dno
[PULLEYS ‘Suoj oqn] 12212 “9|ISSIS
Ápsow ‘ueu 0] [ “JURIBRLY ‘aM
‘эле ‘orydsowounoe *prongioueg
"ugreqjs juaue
-[ Әәл) *feouipur&oqns dno [eururejs
‘Buoy eqn) 'sno[npuod *oje[[oorpod
*snodouinu Зиви8елу Jou “иәэ18 *[[euis
+ corydsowounoe чеүпчпу-шлодәпип y
‘areur[92p pue 2по|
sjuaure[y әәлу *aqni Əy} шолу paxop
-әр ‘adie, = dno jeurweys ioys
eqn) :әјешүәәр “3|ISSISGNS “IZ
“JURIBRIY JOU ‘1M ‘a31| *ongdaoui02 47
‘PaAINOUL pue uoys juour
-P[tj әәлр шим Uso peduijs *93.P]
dno ¡eurure]s "Juo| x eqn) ‘spu
-I[99P 10 [PJUOZLIOY *ajej[22rpad ој
3[rssosqns *([-z “JURIBRIJ *wo[[a4 10
3]« “эвлер omdaourourjoe "prongaoueq
'pexinour pue poys јиәше[у
331] TUTYIIM пәәл# рэа pue 93.18]
dno [eururejs ‘Suoj əqn} дәәлә(дп)
“Q[ISSOS “JUPIBRIY “MOT[OA 10 3]I[«
9318] *orgdaourourjoe *prorgaoueq
juasqy
juasa4q
juaso4q
]u3s214
juasqy
EOLI3UIPOS
AJA `вәгрир ISAM “SA AS
ш 000€ әлоде ‘nəd
ѕиоцеләјә цац о]
pru je Jopenoy pue плә,]
ѕиоцелә[ә ущ
0] MOJ I? səpuy penu
suone4o[9 MOT те ‘nd
AN pu? Iopenoq MS
0с “Bo
c
S1]]DIOUGULA H
uossoul
-0uajsopnosq “BQNS auauls
d `q I
Dua] "oqns suaus]
əuəus] "qns ousus]
u0114201d7]
Jequinu
эшоѕошоц?у
роо paas uo
uP[auio]4uqd
әүпәо]
iad səma
(Zojoydiow [P10] 4
wa]sopnasd
3p]e3uo[
uornqugst(]
satdads
Јо 19QUINN]
snuazqns
10 snug
‘areapı[[P2ouauAH aqua 3eaoPpr[[Areury јо eıauadqns pue рләпәЗ ay) jo uosureduio")
E "MEL
403
—
—
Phylogeny of Tribe Hymenocallideae
Meerow et al.
"|Р 19 моләәр
‘|B 19 MOLI
в ә моләәрү
"JE 19 MOII9[A
үе 19 MOLI]
qe Ja MOLI],
0801 1PAV
ye 19 MOLI9 N
Je 19 моләә[\|
ye 19 MOLI ÍA
“pe 19 MO.139 [A
JE jo мого
(6661) [e 19 моләәрү
(6661) [e 19 моләәң
(6661) e 19 моләәу
(6661) [E 19 *o129[N
—
(6661) "T€ 19 MOI99]A
OLOLITAY
(6661) 18 19 Mo132]|N
(6661) TF 19 моләәрү
(6661) 1° 19 моләәр
(6661) `I? 19 моләәрұ
—
(6661) 18 19 Mo12929[N
(6661) ЧЕ 19 Mo1229[N
S20LLEAV
(6661) 18 19 M0129]|A
7.
jdeyg pumantad ouquniumg
ATRAS (qlo H) saisusopmb uo poda]
1
MOIDI Y NLA (әрїеүәд) 118D24Da
M0122 N
(әрлејәд) иохғәшоиәјѕорпәѕ "апе ouauis]
моләәү C[pur]) руруәйї8ио] `]
MOLVA СӘҢ) очокту 7845 ouaus]
‘boe f vaopfissioanu `]
моләәр Y пвәлә=) (SPBIRA) NSƏYNDY *]
«Моң (ARG y ZINY) sonoupum suaus]
“SUBS эмо ^H
"Neg, vyofipuvyonə ^H
‘moo (TTA) 2170/70 y
"WOY “JA Pond)? Y
199MS (Чөң) pipofimop si]]roouaui&g
Joords 4-qu4]
3u93 “ju
ПОР ѕпотләла J0 ‘OU UOTSSIIIR que gus)
похве,
Volume 89, Number 3
2002
`рәїкәїриї ISIMISUYJO ss9[un HLA ye pausodop 318 SIIYONOA ПУ тгәвәрццүвәоцәшАн jo savduanbas VNC 19] sI9qunu UO0ISSIIIB que gus) Mou pue SIIYINOA y 9149],
404
Annals of the
Missouri Botanical Garden
Table 3.
Characters and character states used in the cladistic analyses of Hymenocallideae based on morphology.
Character
States and coding
1. Elongate pseudostem
. Flower number
3. Flowers sessile/pe dicellate
7. Ponant — pancra
erianth symme
10.
11. Staminal cup shape
12. Staminal с es striping
13. Free filame
14. Free —
15. Pollen grain size
16. Pollen grain
17. Exine reticulum
18. Ovules per locule
19. Seed per locule
20. Phytomelan on testa
n
Ф
21. Seed cc
22. Saad shape
23. Most common diploid chromosome number
erect
absent
2-10 (0); solitary (1); >10 (2)
sessile (0); pedicellate (1)
0); declinate/horizontal (1); pendent (2)
shorter than tepals (0); longer than or equal to tepals (1)
straight (0); curved (1)
—
20 ( )); 1
numerous (0); 2-5 (1); 1 (2)
esent (0); absent (1)
oat not fleshy (0); fleshy (1)
P € (0); globose (1)
46 (0); :
p
0); present (1)
atioid (0); funnelform-tubular (1); + funnelform (2
—
actinomorphic (0); zygomorphic (1)
white (0); vellow (1); green (2)
present (0);
absent
(1)
rotate or funnelform (0); cylindrical (1)
present (0); absent (1)
incurved (0); straight (1); declinate (2)
longer than cup (0); shorter than cup (1)
very large (0); large (1);
auriculate 9 not т
medium (2)
е (0); n
—
edium (1
6:00. " 2-10 (2); 2—4 (3)
34 (1); 46, 40 (2)
SEQUENCE ALIGNMENTS
Both sequence matrices were readily aligned
manually using the program Sequencher (Gene-
Codes, Inc., Ann Arbor,
needed to be inserted.
Michigan) as few gaps
CLADISTIC ANALYSES
Pamianthe (tribe Clinantheae) was used as out-
group for all analyses. In larger sequence analyses
Table 4.
morphisms: +
= (0,1); * = (0,1,2).
(Meerow et al., 1999, 2000a), this genus resolves
as most closely related to the Hymenocallideae. Pa-
mianthe and Leptochiton (the latter putatively the
least derived genus in the Hymenocallideae; see
discussion below) share two four-base sequence el-
ements in the trnL-F region (bp325-328, 821-824)
that are absent from the rest of the Hymenocalli-
deae. Phylogenetic analyses were run using PAUP*
version 4.0b8 beta (Swofford, 1998). An exhaustive
search of all possible tree topologies was conducted
Character state matrix for cladistic analysis of 23 morphological characters in Hymenocallideae. Poly-
Taxon
Matrix
Character
Hymenocallis acutifolia
Hymenoc allis eucharidifolia
Hymenoc allis tubiflora
smene amancaes
Ismene hawkesii
Ismene longipetala
Ismene narc —
Ismene vargasii
Leptochiton —
Pamianthe peruviana
1 2
12345678901234567890123
00001000000110000221112
0*001000000130000221112
0*001000000130000221112
0*001000000130000221112
0*001000000130000221112
10111+00100001000321110
10111+00000001000321110
10110021011120211321110
10111+00000001000321110
10120110211110211321110
01001100+00001000110111
10111000000001110000000
Volume 89, Number 3
2002
Meerow et al.
Phylogeny of Tribe Hymenocallideae
for trnL-F. For ITS, the morphological, and all com-
bined analyses, branch and bound searches were
conducted. Support for internal nodes of the trees
was determined with 5000 replicates of branch and
1985) and by
calculation of Bremer (1988) decay indices (DI) us-
1999). A
branch and bound search was implemented for
bound bootstrapping (Felsenstein,
ing the program TreeRot (Sorenson,
each constraint statement postulated by TreeRot. A
bootstrap value of 50-64% was considered weak,
65-74% moderate, and 75-100% strong support.
Combining independent character matrices,
whether both molecular or molecular and morpho-
logical, very often increases the resolution of the
ingroup and the bootstrap support of the internal
nodes of the phylogenetic trees (Olmstead &
Sweere, 1994; Chase et al., 1995; Yukawa et al..
1996; Rudall et al.. 1998; Soltis et al.. 1998: Mee-
row et al., 1999), Nonetheless, there is controversy
about whether different data sets should be ana-
lyzed separately or together (De Queiroz et al.,
1995 1996). Congruence of the
independent matrices has generally been demon-
; Huelsenbeck et al.,
strated before they are combined, but it has also
been argued that incongruence should not be a pre-
determined factor against doing so (Dubuisson et
al. 1998; Seelanan et al.. 1997). Miyamoto and
Fitch (1995) argued that data sets should always
be analyzed independently, as underlying assump-
tions, constraints, or weighting strategies will vary
from data set to data set. Kluge (1989) and Nixon
and Carpenter (1996) argued that simultaneous
analysis of multiple data sets better maximizes par-
simony and allows secondary signals to appear from
the combined data. Bull et al. (1993), Rodrigo et
al. (1993), and De Queiroz (1993) advocated com-
bining data only after a statistical test of congru-
ence, what Huelsenbeck et al. (1996) called “
ditional combination."
con-
Before combining the data
sels, we performed a partition homogeneity test
(Farris et al.. 1994, 1995) on the variously com-
bined matrices, using a branch and bound search.
RESULTS
MORPHOLOGICAL MATRIX
With all characters unordered, two most parsi-
monious trees (Fig. LA, one shown) were found of
length = 37, consistency index (CI) = 0.86, and
retention index (RI) = 0.88. Sixteen of the 23 char-
acters used were parsimony informative. In both
ees, Hymenocallis is monophyletic (bootstrap =
ш. DI = 1). while /smene is paraphyletic. /smene
longipetala (subg. Elisena) and I vargast (subg.
Pseudostenomesson) are sisters in both trees. Lep-
tochiton is sister to both Hymenocallis and Ismene
in one tree (Fig. ТА). The 6 apomorphies at the
ancestral node are an increase in pollen grain size,
auriculate pollen grains, reduction in ovule number
from more than 20 to 16 to 20; reduction in number
of seeds per locule; and evolution of globose, fleshy
seeds. Apomorphies for Hymenocallis (Vig. ТА) are
the absence of an elongate pseudostem, predomi-
i erect flowers, and 2n = 46, 40
Other than Hymenocallis, the only
clade with strong bootstrap support is the sister re-
nantly sessile an
chromosomes.
lationship of /smene subg. Elisena and subgenus
Pseudostenomesson (100%, DI = 6).
apomorphies: perigone tube length reduction, non-
based on 7
fra-
grance, cylindrical staminal cup, and smaller non-
pancratioid floral morphology. loss of floral
auriculate pollen grains with less coarse exine
reticulum. If all of the characters are ordered as
irreversible, a single tree is found of length = 48,
CI = 0.67 and RI = 0.88 (Fig. 1B). There is
moderate bootstrap support for a monophyletic Zs-
mene (65%: DI = 2; apomorphies: elongate pseu-
dostem, pedicellate and declinate/horizontal flow-
ers, and 2—4 ovules per
with C
locule). There is weak
support for the sister relationship of Hymenocallis
and /smene (56%. DI
in ovule and seed number, respectively: and the
1; apomorphies: reduction
loss of phytomelan from the testa). Leptochiton is
moderately supported as sister to both (65%, DI =
1: apomorphies: reduction in ovule and seed num-
ber and the evolution of a fleshy seed). /smene subg.
Ismene has a 91% bootstrap and DI = 2. /smene
subg. Elisena (1. longipetala) and subgenus Pseu-
aa (I. vargasit) are again sister groups
1 100% bootstrap and a DI = 9. A monophyletic
ola receives 87% bootstrap support with
DI = 4. Hymenocallis latifolia, H. glauca, and
Н. eucharidifolia form a monophyletic group with
^
>
bootstrap support of 60 and DI = 1. This same tree
topology (Fig. 1B) is 40 steps long with CI = 0.80
and RI
with the topology as a constraint with all characters
= 0.80 if a branch and bound search is run
unordered.
PLASTID trnL-F SEQUENCES
Using trnL-F sequences, which provide 7 parsi-
mony-informative base substitutions, three equally
most parsimonious trees are found of length = 82,
CI = 0.99, and RI = 0.88 (Fig. 2. one tree shown).
All three trees resolve a monophyletic /smene with
81% bootstrap support (DI = 2), and Leptochiton
as sister to the rest of the tribe but without support.
A monophyletic Hymenocallis is resolved as sister
to /smene in one tree (Fig. 2). but Hymenocallis and
Annals of the
406
Missouri Botanical Garden
(€ 314BL 398) qougaq ey} BUO[E sarqdıouode әле SIYIURIG MO| 94 s1oquainu pue soul] [E21LI9A
`ѕәѕәцјиәлва ut JIP (ор) зәори! Арәәр pue sa3ejuaniad de.nsi00g "sy13ua]| Youelg 318 SIYIUPIG gaoge SI9QUINN "9[QUS19A9.LIL se p2312p4o 31E SLIJIBIBYO Пе J! punoj 331] snoruouirsaed
вош a[3uIŞ “> `рәләрдоип ѕләЈәвЈецо ПЕ Чим punoj S331} snoruoursıed 150Шш OM] jo 910 'V— “SIIPOBIBYO ¡eor3o]oyd.1ow uo pəseq әвәрү[еәоиәшАң 10] ѕшеі8ореүэ
eueiAnjed ayjueweg
sisueojinb uojiu2ojde1
nseBsen eueuis|
еүезәа!биој eueuis|
EJO|JISSIDJ8U Guauis|
пѕәумец eueuis|
S9PIUBUJE Guauis|
EIJOJI]8| SIJJEÉ2OuOUIÁAH
eone/6 sijesouawAyY
eyopjipueyona ѕ=цүвэоиәшАн
EJ10/J1GN] sijje20uauiAH
еојцпәе sije2oueuiAH
L1 ‘91 'SL'EL
sdas
ё Sppe £z JapeJeyo
19 ebueyo
lr '/8)$
02 ‘61 ‘81
U `95)Е
eueiAniad ayjuejwed
sisusoynb uojiu2ojde
BIOYISSIIIEU ӘШӘШ$®]
пѕәумец әиәшѕ]
ѕәвэиеше әиәшѕ]
ııseBıen әиәш8]
ејејәаібиој eueuis|
еуојцеј ѕүеэоиәшАн
ganejD =вэоиәшАн
EJJOJIPUEYINA sijjesouawAY
&JojJIqn) sij[e2ouauiAH
Boyne SIJ|B2O UOUIÁH
L1 '9V'SIE
‘LL 'OL'Z'S
d əmi
Meerow 407
Volume 89, Number 3 al.
Nu of Tribe Hymenocallideae
Ismene narcissiflora
Ismene longipetala
Ismene vargasii
Hymenocallis eucharidifolia
2
20 " og. ge
— — Hymenocallis latifolia
7 А „ А
Leptochiton quitoensis
Pamianthe peruviana
Figure 2. One of the three most parsimonious trees found by cladistic analysis of plastid trnL-F DNA sequences
for the тЫ ‘allideae. Numbers above branches are branch lengths; numbers below branches are bootstrap per-
centages, follow
all three trees.
Ismene form a clade in all three (73% bootstrap, DI
= 1). Ismene subg. Ismene (1. narcissiflora) and Eli-
sena (1. longipetala) are resolved as sister groups
in all three trees with a bootstrap of 70% (DI = 1).
ITS SEQUENCES
ITS provides 50 parsimony. informative charac-
ters, and 9 trees of length = 209, C 0.73. and
RI = 0.77 were found (Fig. : 3). In all o of the trees,
Leptochiton is resolved as sister to both Hymeno-
callis and Ismene (Fig. 3A), but without significant
support. jede allis is monophyletic (bootstrap —
97%. DI = 5). but /smene is monophyletic in only
2 of the 9 trees (Fig. 3B, one shown). However,
Ismene subg. Ismene (1. amancaes, 1. hawkesit, 1.
narcissiflora) is monophyletic with weak bootstrap
support (59%) and DI = 1 (Fig. 3B).
ed by decay indices (italic). The large arrow indicates a node that collapses in the strict consensus of
COMBINED £rnrl.-F AND ITS SEQUENCES
The P value from the partition homogeneity test
= 0.93. indicating that the trnL-F and ITS se-
quence matrices were highly congruent. Six most
р arsimonious trees were found of length = 292, СІ
= 0.92, and RI = 0.77 (Fig. 4). In all trees, Hy-
menocallis and Ismene are monophyletic sister gen-
era with bootstrap support of 94% and a DI = 3.
Leptochiton is sister to both, but without significant
support. Bootstrap support for a monophyletic Hy-
menocallis is 98% (DI = 5), but only 68% (DI =
1) for a monophyletic /smene. The only other inter-
nal resolution within /smene that receives bootstrap
support is a sister relationship between /. narcis-
siflora and I. hawkesii (both within subg. Ismene) at
84% with DI =
Annals of the
408
Missouri Botanical Garden
"(9i[e11) вәә1їрш 4вәәр Aq рәмо[үо} sedejuao1ad 21151004 318 ѕәцәиві ^o[oq ѕләдшпи :sy¡3ua] Y9UBIG әле SIYIUP.IG әлофе sroquinv 'sioouawukg 0] dno13 191818 one[4udouour e
SI 2uauis| YOTYM UL S221 OM] JO эц) “g— “SIAM snoruourrsed. A([enbo әши jo snsuasuoo LYS “y — ‘seouenbes GJ] YN Qu uo paseq arapı[[e2ouauAH 10} sweisopey) сє 310314
euegejaniad әціиешед
sisueojinb uojiu2ojde]
eyojipueyonea $=јеэоиәш/Ан
eanejD si[[R2ouauiÁH
&JOJJIqn] $1јеэоиәш/Ан
EIJOJI1NIE sij[RB20uGUIÁH
EI/OJ1J8] ѕ=1івэоиәш/Ан
пѕебіел ououus /
ejejedibuo| eueuis|
saeouewe ououus /
lisaymey eueuis|
BlOJJISsI2.eu eueuis|
V
BuelAniad aujuejuegd
sisuaojinb uojiu2oojde]
IISEDJ8A ououus /
eٳfoJpıeyone si[j[eR20uauJAH
ganejD si[jR2ouauiAH H
EIOMOMD si[RB2ououlAH
8IJ0JI]n2P $1јеэоиәш/н
вое Sij|EB20uGulÁH
ejejadiBuo] eueuis|
seeouewe әиәш]
пѕәҳмец әиәшѕ]
EJO|JISSIDIBU ououus /
Volume 89, Number 3
2002
Meerow et al.
Phylogeny of Tribe Hymenocallideae
409
Figure 4.
ITS sequences. Numbers above branches are branch lengths: numbers below branches are bootstrap percentages fol-
lowed by decay indices (italic). T
COMBINED SEQUENCE AND
MORPHOLOGICAL MATRICES
The P value of the partition homogeneity test was
0.0003,
tween the morphological and DNA sequence data
indicating significant incongruence be-
matrices. Much of the apparent incongruence can
be attributed to the weak resolution of the morpho-
logically based topologies, and we felt that it would
still be informative to combine the two matrices in
a single analysis. Of the 1565 characters included,
76 were parsimony informative. A single tree was
found of length = 332, СІ = 0.92, and RI = 0.79
(Fig. 5A). Hymenocallis is monophyletic with 100%
bootstrap support (DI = 8), but /smene is paraphy-
Ismene narcissiflora
Ismene hawkesii
Ismene amancaes
Ismene vargasii
Hymenocallis glauca
Hymenocallis eucharidifolia
Pamianthe peruviana
PERU
PERU
PERU
Ismene longipetala ecuanor, PERU
PERU
Hymenocallis latifolia FLORIDA. UPPER ANTILLES
Hymenocallis acutifolia mexico
Hymenocallis tubiflora NORTHERN souTH AMERICA
MEXICO
MEXICO
Leptochiton quitoensis SW ECUADOR, NW PERU
BOLIVIA, PERU
One of six most parsimonious trees found by cladistic analysis of combined plastid trnL-F and nrDNA
). The larger arrow indicates a node that collapses in the strict consensus of all six trees.
letic. Bootstrap support for the monophyly of Ismene
(DI = 2),
subg. Elisena (1. longipetala) and Pseudostenomes-
subg. /smene rises to 81% but /smene
son (d. vargastt) are sister groups — = 97%,
DI =
= 1) as sister to Hymenocallis. Leptoc — Is again
I) weakly supported (bootstrap = 57%, DI
sister to the other members of Hymenocallideae but
without support. If trees one step longer were also
retained in the search, in addition to the single
shortest tree (Fig. 5A). a single, fully resolved tree
333. CI = 0.90, and RI = 0.77 was
found (Fig. 5B). In this tree (Fig. 5B), both Hymen-
ocallis and Ismene are monophyletic sister genera,
of length =
as are Ismene subg. Elisena and Pseudostenomes-
son.
Annals of the
410
Missouri Botanical Garden
"suua| qoueaq 318 SIYIURIG JAOQB SIIQUIN y `рэч!в191 318 Vc
ut painjord 331] uey} 19200] days 3UO Saa] jl punoj 3341] [tuo ippe әш E ° (оен) 839)IpUI \вдәр íq рэмо[[0ј sogequa21ad 421181004 ӘЛР SIYIURIG MO[3q s19qunu "sy 1Suo] youe1q
ӘЛЕ SIYIUPIG 9AOqE SI9QUNN *331] snoruouits ed jsou әш "W— "SJOLI]EUI әәиәпһәѕ YNA pue [eotsojoydiouw p2urquioo uo peseq әрәреәоиәшАң 10] ѕшрізореүэ 6 3131
g V
euelAniad ayjueiwed euejansad ayjuejwed
sisueogjinb uodu20jde
| sisuaoynb uo}y901de7
еџоџриецэпә sjjje2ouaw AH EIJOJIPIIEYINO sijjesouawAyY
гопејб sijje2ouaw AY eonejb sjjje2ouaw AY
=
6S
елордап1 sijje20ouauiAH ds BIOJJIqn] SJjje2oua ui AH
95
EIJOJI jn 2B siJj[B20ua ui AH eojnnoe Się20uə WÁH
Boye =еэоиәш/Ан BIJOJI18] SiJ&e2ouaurAH
¡sebiea auaws] ysebien әйәш$|
ejejedibuo| eueuis| ejejadibuo] euauis|
ѕәвэиеше әиәш] saesguewe auauis[
пѕәҳмец әиәшѕ] пѕәҳмец әиәш] 2
210115812Јви BUGIS] EJO|JISSIDJBU ӘШӘШ$|
zz
LL
6
oL
ez
$
8
y zu
lz
ц
6z
LL
L
6
t
zu
Volume 89, Number 3 Meerow et al. 411
2002 Phylogeny of Tribe Hymenocallideae
DISCUSSION the non-fleshy seeded Andean endemic Clinan-
Both plastid (Meerow et al., 1999) and ITS (Mee-
row et al., 2000a) sequences strongly support the
position of the tribe Hymenocallideae as a mono-
phyletic group within the Andean tetraploid clade
of the endemic American Amaryllidaceae that i
sister to the newly recognized tribe Clinantheae
Meerow (Meerow et al., 2000a).
Clinantheae are. uniformly
The seeds of the
flat, winged, and
with phytomelanous testas. There are links between
Leptochiton and Pamianthe that Stapf implicitly
recognized, most notably the plesiomorphic pres-
ence of phytomelan in the testa of Leptochiton’s
seed [of which Meerow & Dehgan (1985) were un-
aware], but also the numerous ovules of this genus
(plesiomorphic as well). In the ITS phylogeny pre-
sented by Meerow et al. (2000a), support for Pa-
mianthe as sister to the rest of Clinantheae (vs. a
dry,
sister group relationship to Hymenocallideae or an
unresolved position) was considerably weaker when
the aligned matrix was not successively weighted.
This is not surprising given that both genera occupy
a basal phylogenetic position in their respective
clades herein.
The difficulty of relying on morphological char-
acters alone to generate phylogenies in Amarylli-
daceae has been discussed (Meerow, 1995; Meerow
20005).
for many morphological characters in the family.
et al., given a high degree of homoplasy
Our analysis (Fig. 1) generates trees with relatively
high CI and RI, but parsimony is still not able to
resolve /smene nor consistently place Leptochiton in
the basal position within the tribe with unordered
morphological characters alone, in contrast to se-
quence data (Figs. 2-4), which also provide (in the
combined trnL-F and ITS matrix). over three times
the number of phylogenetically informative char-
acters of morphology alone. The combined plastid
and nuclear sequence matrix produces the most ful-
ly resolved shortest trees. To “force” this topology
upon any of the other conflicting data matrices re-
quires either ordering characters or accepting lon-
ger trees (albeit only one step longer in the com-
bined sequence and morphological analysis).
When biogeographic information is optimized
upon the combined plastid and nrDNA tree (Fig.
4), the gene phylogeny supports an origin for the
tribe in the central Andes, inarguably a locus of
diversity for the Andean tetraploid clade of the
2000a). with a vi-
cariance event that gave rise to the largely North
American Hymenocallis. Leptochiton, with 16 to 20
Amaryllidaceae (Meerow et al.,
ovules per locule and a phytomelanous testa, oc-
cupies a relict position in the tribe with links to
theae. However, it is the genus /smene that reflects
the patterns of floral morphological diversity that
occur in the Eustephieae, Clinantheae, and Steno-
messeae (sensu Meerow et al., 2000a). /smene subg.
Ismene retains the plesiomorphic pancratioid floral
morphology of Leptochiton, Pamianthe, and Hymen-
ocallis, while the smaller /smene subg. Elisena and
subg. Pseudostenomesson express floral novelties.
Ismene subg. Pseudostenomesson, occurring at the
highest elevations of any member of the tribe, might
be the youngest element of the polymorphic /smene.
since the Andes likely did not extend above 1000
m elevation before the Pliocene (10 MY BP: Van der
Hammen, 1974, 1979). Analogous patterns of floral
diversity are found throughout the tetraploid An-
dean clade of the American Amaryllidaceae. In the
Clinantheae, the low- to mid-elevation genera Pa-
mianthe and Paramongaia Velarde have pancra-
tioid floral morphology, while the mostly high-ele-
Clinanthus Herb. putatively
ornithophilous flowers. In the more distantly related
vation has colorful,
petiolate-leafed Stenomesseae, Eucharis has the
pancratioid flower; Plagiolirion resembles a mini-
ature /smene subg. Elisena; and Stenomesson and
Urceolina exhibit colorful, putatively ornithophilous
flowers. Finally, in the Eustephieae, which is sister
to rest of the Andean clade (Meerow et al.. 2000a).
the full range of variation is evident in a single
genus. Hieronymiella Pax (Hunziker, 1969).
recurrent pattern suggests a scenario of rapid mo-
1984) within this
1987).
The relatively low number of phylogenetically in-
This
sale evolution (sensu Stebbins,
monophyletic, tetraploid group (Meerow,
formative base substitutions in our sequence anal-
50 for ITS)
supports a hypothesis of a relatively recent radia-
tion within the Hymenocallideae tied to the rise of
the Andes. This seems most significant relative to
yses of non-coding regions (7 for trnL-F;
Ismene, the most polymorphic of the three hymen-
ocallid genera, and the only one that has adapted
to high elevation.
Hymenocallis is most speciose in Mexico (Bauml,
1979). with a secondary area of diversity in the
southeastern United States (Smith & Flory, 1990,
2001: Smith et al.. 2001). Only
species have been reported from South America:
the broadly and coastally distributed H. littoralis,
H. pedalis, and H. tubiflora. The genus does not
occur at all in the Andes, and H. tubiflora is the
three described
only species of the three that is restricted to north-
South America (including Trinidad-Tobago).
The known distribution. of
егп
the Hymenocallideae
suggests two possible hypotheses, either a long-dis-
—
ance dispersal event from the Andean center of
412
Annals of th
Missouri ы Garden
origin, or extinction of intervening populations of a
proto-Hymenocallis ancestor. The fleshy seed of Hy-
menocallis is the largest of all the endemic Amer-
ican Amaryllidaceae, exhibits no dormancy, and
germinates within 3—4 weeks after release, whether
substrate (Whitehead & Brown, 1940;
obs.). The relatively heavy seed does not im-
or not in
pers.
mediately seem amenable to long-distance dispers-
al, and no dispersal agent other than water has even
been suggested for the genus. Thus ancestral ex-
tinction is a more convincing hypothesis, but with-
out a better understanding of the historical bioge-
and a well-resolved
ography of Hymenocallis
-
phylogeny of the genus a likely explanation for its
distribution cannot be determined.
In summary, combined trnL-F and ITS sequenc-
es support the Meerow and Snijman (1998) treat-
ment of Hymenocallideae with three genera: Hy-
menocallis, Ismene, and Leptochiton. Leptochiton is
sister to the Hymenocallis/Ismene clade and retains
two plesiomorphic characters of the Andean tetra-
ploid clade: 16 to 20 ovules per locule and a phy-
tomelanous seed coat. The central Andean ende-
mism of /smene and Leptochiton and the absence of
Hymenocallis from this region further suggest a vi-
cariance event at some point subsequent to the or-
igin of the tribe. It is thus appropriate to formalize
the recognition of the two new subgeneric combi-
nations within /smene.
Ismene subg. Elisena (Herbert) Meerow, comb.
nov. Basionym: Elisena Herb.,
ceae, 75, 201. 1837. TYPE: /smene ringens
(Ruiz & Pav.) Gereau & Meerow, Novon 3: 29.
1993.
Ismene subg.
Amaryllida-
Pseudostenomesson (Velarde)
Meerow, comb nov. Basionym: Pseudosteno-
messon Velarde. Rev. Cienc. (Lima) 51: 47—51
1949. TYPE: /smene vargasii (Velarde) Gereau
& Meerow, in L. Brako & J. — Monogr.
Syst. Bot. Missouri Bot. Gard. 45: 1253. 1993.
Literature Cited
Baker, J. G. 1888. Handbook of the Amaryllideae. (
Bell and Sons, London.
Bauml, J. A. 1979. A Study of —
ed in Mexic
versity, Ithaca, New Yor
Bremer, К. 1988. The limi of amino acid sequence data
in angiosperm phylogenetic reconstruction. Evolution
42: 198-213.
Bull, J. J., J. P. Huelsenbeck,
Swofford & P. J. Waddell. T 'urtitioning — com-
ining — in phylogenetic analysis. Syst. Biol. 42:
384-397
C — M. We D. W. Stevenson, P. Wilkin € P. J. Rudall.
995. Monocot systematics: A combined analysis. Pp.
»eorge
V AN
Genus Hymenocallis
. Thesis, Cornell Uni-
. Cunningham, D. L.
ex
685-730 in P. J. Rudall, P. J. Cribb, D. F. Cutler & C.
J. Humphries (editors), Monocotyledons: —
and Evolution, Vol. 2. Roya ‚ Kew
кзы n, R. M. T., H. T. Clifford & P. E ‚ 1985. Тһе
amilies of the Monocotyledons. ее Вег-
—
ЖО, L^ d 1993. For consensus (sometimes). Syst.
Biol. + 68-372.
moms & J. Kim. 1995, Separate versus
combined analysis of phylogenetic evidence. Ann. Rev.
Ecol. Syst. 26: € э ›81.
Douzery, J. P., А. ни оп, ДЕ Когеѕ, Н. Kurzweil, Р.
Linder & M Chase: Э. Molecular ibe ties
of Diseae ос ra contribution es nuc ae
ribosomal ITS sequences. Amer. J. Bo E
Dubuisson, ebant-Mauri & y ps ier. 1998.
Molecules m alan: Conflicts a congruence
ia the fern genus Tric — ا eae)
Molec. nel. Evol. 9: 390—X
"arris, J. S., A luge & E Bult 1994,
nificance of incongruence. Cladistics 10: :
— ———, M. Kiillersjó, A. G. Kluge & C. Bult. 1995. Con-
struc ang a significance test for incongruence. Syst. Biol.
44: 570—572.
Felsenstein, J. 1985. Confidence limits on а таран
An approach using the bootstrap. Evolution 39: 783—
791.
=
Testing sig-
5-319.
Flory, W. S. 1976. Distribution, chromosome number and
types of e species of Hymenocallis. The Nucleus
19: 204—2
Gereau, R. E. A. W. Meerow & L. Brako. 1993. New
Combinations in Hippeastrum, Ismene, and Leptoc hiton
(Amaryllidac eae) for the Flora of Peru. Novon 3: :
Grant, V. 1983. The systematic and — cal distri-
— жр in the temperate North
n flora. Bot. 144: 439-449,
Hammen, E Van der on The Pleistocene changes of
— and climate in tropical South America. J.
Biogeogr. 1: 2-26.
179. History of the flora, vegetation and climate
in the C Vienn Cordillera Oriental during the
million years. Pp. 25-32 in H. Larsen & L. B.
Tropical Botany. Academic
ast five
Holm-
Nielsen (editors), Press,
1837.
ê andi
Huelsenbeck, | P. J. J. Bull & €. W. Cunningham. 1996,
ues. data in кин tic analysis. Trends Ecol.
Evol. | 52-15
Hunziker, A. un Estudios sobre Amaryllidaceae. III.
Sinopsis Dilanê) de Hieronymiella, y — Ar-
gentinas юше о Kurtziana 5: 343-3
Hutchinson, J. 4. Families of — Plants. n 2.
Monoc — Ist ed. MacMillan, London.
. 1959. Families of Flowering Plants, Vol. 2. Mono-
cotyle dex 2nd ed. Clarendon Press, Oxford, U.K.
Kluge, A. G. 1989 phylo-
ic hypot hod of а among FEpicrates
(Boidae, Se ш! Syst. Zool. -25.
Meerow, A. W. 1987. The — ne systematic rela-
tionships of о — h ye e Baker
ано ў :eae). Taxon 36: 566-5
Amaryllidaceae. James Ridgeway and
. A concern for evidence and a
8
=
9. Systematics of the — lilies, Eucharis
and culpan ннн Ann. Missouri Bot.
Gard. 76: 136-22
=
Volume 89, Number 3
2002
Meerow et al. 413
Phylogeny of Tribe Hymenocallideae
————. 1990. 202, ише In G. Harling (edi-
tor), Flora of Ecuador No. 41. University of Göteborg.
————. 1995. Towards a — ny of the —
169-179 in P. J. Rudall, P. J. Cribb,
. J. Humphries (editors), Monoc —
Systematic 5 үе Evolution, Vol. 2. Royal Botanic Gar-
de n5,
—
^
é Jehgan. 1985. The auriculate pollen grain
of Hence кл quitoensis Herb. (Amaryllidaceae) and
its systematic in eee mer. J. Bot. 72: 540—5:
Sehr: D. A. Snijman. 1998. alii Pp. 83—
ы їп К. Kubitzki (editor),
Vascular Plants, Vol. 3. Springer- — zerlin.
——, М. К. Fay, €. L. Guy, Q-B. Li, К. Q. Zaman &
hase. 1999. —— er maryllidaceae
based on cladistic analysis of plastid rhel. and trnL-F
sequence data. Amer. J. hey 86: 1325-1345
————, C. L. Guy, Q. Li & S-L. Yang. 2000a. Реза
of the потем Pese eae i based on nrDNA ITS
sequences. Syst. Bot. 25: 70
———, M. Кау, M. W. Chase, С. E Q. Li, D. Snij-
man & S-L. Yang. 2000b. Bis f ‘the Amarylli-
daceae: Molecules and morphology. Pp. 308-382 in K.
Ф
~
Families and Genera of
е»
5
Wilson & D. Wallace (editors), Мох cols: Systematics
ind Evolution. CSIRO Press, Sydne
—— M. M. &W.M. h. 1995. Testing species
phy logenies and phylogenetic methods with congru-
14: 61—76.
1965. The experimental taxonomy of the
est African — Рапат L. (Amaryllida-
eae). Kew Bull. 19: : 4
Müller- Doblies, D. & U. мл. Doblies. 1996. Tribes and
subtribes 2 some species combinations in Amarylli-
daceae J. St.-Hil. emend. R. Dahlgren et al. 1985. Fed-
des Rep. 107070 S.c.
Caen de
sis. р — s 12: —24
. А. —
in — netic systematic
using scalps molecular е sets in the Solanaceae, Syst.
Biol. 43: 467—481.
Pax, F. id
In A. Engler
Pflanzenfamilien Il. Teil 5:
mann, Leipzig.
р)
= On simultaneous
— Combining data
— Ап empirical approach
lasted. С. &
Amary llidoide 'ae-Narcisseae-Eucharidinae.
Prantl (editors). Die natürlichen
110-11. Wilhelm Engel-
~
Ravenna. P. 1980. A new yellow-flowered —
(Amaryllid ) from north Peru. Bot. Not. 133:
98.
Rodrigo, A. ( . Kelly-Borges, P. R. Bergquist & P. L.
Bergquist. ed A randomization test of the null hy-
pothesis that two cladograms are sample estimates of a
— phylogenetic tree. New Zealand J Bot. 31:
a 257-26 к.
а PJ, Chase, D. F. Cutler, J. Rusby &
. De Кы 1998, ——— M de * sys-
tematics of Asteliac eae and Hypoxidaceae. Bot. J. Linn.
Soc
‚ 127: 1—
Salisbury, R. A. . On the a of rare plants.
Trans. Hort. з sh lon 1: 261—3€
Sealy, J. R. 1937. Lepidochiton — nsis. Bot. Mag. t.
949]
Review of the genus Hymenocallis. Kew
— Bull. pa 201-240,
Seelanan, T., A. Schnabel & J. Wendel. 1997. Congruence
and consensus in the cotton tribe (Malvaceae). Syst.
Bot. 22: 259-2
Smith, G. L. & W. 5, Flory. 1990. Studies on — allis
henryae (A 3rittonia 42:
—
‚ 2001. — — exits eae)
in Texas, with a new varietal combination. Novon 11:
229-232.
—— ———, L. С. Anderson — W. S, Flory. 2001. A New
Species P Pines callis (Am aceae) in Lower
Central — P anhandle exon | |: 233-2
Soltis, D. E., H Soltis E. Mort, M. W. ar
. M. Morton. 1998. Inferring
complex —— s using parsimony: An empirical ap-
pu lh using t Pa DNA data sets for angio-
sperms. Syst. Biol. 47: 3242.
— M. D. 1999. “Tree Hot, version 2. University of
Mic — Boston University, Boston, Massachusetts.
Stapf, O. 1933. Pamianthe peruviana. Bot. Mag. t. 9315.
ие G. L. 1984. Mosai mosaic se —
and angiosperm phylogeny. Bot. J. Linn. Soc. 68: 1
164.
Swofford, D. L.
ENS v. 4.0
с ev oluti tion,
Analysis Using
1998. PAUP: Phylogenetic
) beta. Sinauer, Sutherland, Massac Жы
Taberlet. P.. E. Gielly, С. Pautou & J. 1991. Uni-
versal primers for amplification of three non-coding, re-
gions of chloroplast DNA. РІ. Molec. Biol. 17: 1105-
1110
Traub, H. P. 1962. Ke
species ш Нутепос allis. PL
3
Bouvet.
y to ES aro nera, alliances and
59-72 2.
——— : iM eae.
Plant 3 . Society, La Jolla, California.
—— ———., 1980. Sec tions and alliances. genus Hymenocal-
lis ded ti : 46-49.
;ene А of the American
oO
oN
aan
Life :
Velarde, 1949. Pseudostenomesson un nuevo genero
de iie eee Rev. Cienc. ma) ) 51: 47-5
White, T. J.. T. Bruns, s. Lee & J. Taylor. 1990. T
fication and direct sequenc ом” of fungal ribosomal RNA
genes for phylogenetics. Pp. М. Innis. D.
Gelfand. | Sninsky & T. e ite — PCI
cols: A Guide to Methods and Applications. Academic
Press, 1 F
Whitehead, — OF Brown. 1940. The seed of *
spider ih СМИРАМ Amer. J. Bot.
129-
was ex Н. Ohba, К. M. Cameron & M. Chase.
‚ Chloroplast E phylogeny of Ebr Bendio-
bid (Orchidaceae): Insights from a combined analysis
based on леі, sequences and restriction site variation.
J. PL Res. 109: 169-176
PHYLOGENETIC AND Jun Wen,? Porter P. Lowry IL?
BIOGEOGRAPHIC Jeffrey L. Walck,* and Ki-Oug Yoo’
DIVERSIFICATION IN
OSMORHIZA (APIACEAE)!
ABSTRACT
Osmorhiza Raf. (Apiace ae) consists of 10 species — ‘tly distributed in temperate Asia (1 sp.) and the Americas
(9 s pp.) Osmorhiza berteroi and О. depauperata show an American antitropical disjunction. ДИ North America,
use two species are also disjunctly distributed in eastern pis western North America and the Great Lakes regions.
A phylogene tic analysis was conduc * toc — inter- and — ific relationships based on sequences of the ITS
and 5.85 regions of nrDNA. With Anthriscus, Geocaryum, and Myrrhis as outgroups, Hs monophyly of Osmorhiza is
strongly — d. The ITS she — the basal position of the Asiatic O. aristata and the ‘monophyly of the
nine orld species. The ITS sequence of Osmorhiza aristata is relatively dive ur from those of all other species
even йи it is morphologically similar to the eastern North American О, claytonii and O. longistylis (which form a
clade), suggesting early divergence followed by morphological stasis. Osmorhiza berteroi, O. brac 'hypoda, О. depauperata,
О. mexicana, О, occidentalis, and О, purpurea constitute a monophyletic group (= western North American clade), The
morphologically distinct O. glabrata from the central Andes forms a trichotomy with the eastern North American clade
laytonii and O. longistylis) and the western North American clade in — and maximum likelihood analyses.
The 11 populations studied of the widespread O. berteroi form a clade, and showed little sequence divergence, sug-
gesting recent establishment of the widely disjunct populations following long-distance dispersal. Disjunct populations
of O. depauperata from the Rocky Mountains and eastern North America have an identical ITS profile. Osmorhiza
occidentalis, however, shows a high level of infraspecific sequence dive rgen nce. The ITS phylogeny and ihe low sequence
divergence values suggest rapid рае Яе ‘ation of Osmorhiza in western North America.
Key words: Apiaceae, biogeography, disjunction, morphological stasis, Osmorhiza.
ze
Osmorhiza Raf. (Apiaceae) comprises ten spe- genus (Lowry & Jones, 1984; see also Kartesz &
cies, with one species in Asia, eight species in Ghandi, 1993), the sole Asian species, O. aristata,
North America (three of which also occur in South is closely related to two species found in eastern
America), and one species restricted to the central North America (O. claytonii and O. longistylis). The
Andes. The distribution of species within the genus second distributional pattern is seen in O. berteroi
provides an ideal model for studying the evolution and O. depauperata, both of which exhibit an an-
of both intra- and intercontinental disjunctions, — titropical (often inappropriately referred to as “am-
with three distinct patterns represented among its — phitropical,” see Cox, 1990) disjunction between
temperate North and South America. Also, the more
—
members (Fig. 1; cf. also Lowry & Jones, 1984).
According to the most recent classification of the distantly related O. mexicana subsp. mexicana has
' thank Mark McBroom and Kyoung-Ja Lee for laboratory assistance, and Dan Crawford, Fernando Chiang,
Ric dnd R. Halse, Lawrence Janeway, Jan Jorgensen, Youngdong Kim, Sangtae Lee, Clodomiro Marticorena, Tod Stuess sy,
Kouzi Yonekura, and Shiliang Zhou for help in obtaining le saf material. We are grateful to Bill Burger, Michael Dillon,
and Tod Stuessy for helpful discussions on antitropical disjunctions, Carol Baskin for information on the evolution of
seed dormancy, Steg ге 'n Downie for permission to use sequences of Anthriscus caucalis and Geocaryum macroc arpum,
and three pecado reviewers, David Baum, and editor Victoria C. Hollowell for constructive comments. Curators of
the following herbaria are acknowledged for pe — the examination of their specimens: A, CS, F, ‚ MSC, and
RM. Jeff Walck thanks The Ohio Sti ч University for a postdoctoral fellowship. This study was supported in part by
the Pritzker Laboratory for Molecular Systematics and Evolution of The Field Museum and by grants from the National
Science Foundation (DEB-019605 1 de DEB-0108530) to J. Wen.
* Department of Botany, Field Museum of Natural History, 1400 S Lake Shore Drive, Chicago, Illinois 60605-2496,
U.S.A. wenOfieldmuseum.org.
' Missouri Botanical Garden, P.O. Box 299, St. Louis, sea — 0299, U.S.A.; Laboratoire de Phanérogamie,
Muséum National d'Histoire Naturelle, 16 rue ns 75005 Paris, France. lowry@mobot.org.
' Department of Biology, P.O. Box 60, Middle Tennessee State ‘Unive 'rsit y, Murfreesboro, Tennessee 37132, U.S.A.
jwale kGOmtsu.edu.
' Department of Botany, Field Muse ‘um of Natural History, 1400 S Lake Shore Drive, Chicago, Illinois 60605-2496,
U.S.A. Present address: Division of Life Sciences, College of Natural Sciences, Kangwon National University, Chun-
eheon, Kangwondo 200—701, Korea. а"
ANN. Missouri Bor. GARD. 89: 414—428. 2002.
>
Volume 89, Number 3
2002
Wen et al.
415
Osmorhiza (Apiaceae)
Y
18 >.
Figure 1.—A. Distribution of Osmorhiza (: — eae). —В.
and eastern North American disjunction.
western North American disjunctions.
a range that spans the American cordillera, with
isolated populations extending from northern Mex-
ico to northern Argentina. Finally, a third disjunc-
tion is found in O. berteroi and O. depauperata, in
which populations occur in western North America.
the Great Lakes region, and northeastern North
America.
The disjunct distribution of plants between east-
ern Asia and eastern North America is a classic
topic in biogeography (Gray, 1840, 1878; Li, 1952:
Graham, 1972; Thorne, 1972: Boufford & Spong-
berg. 1983; Wu, 1983: Wen, 1998, 1999, 2001).
and most genera exhibiting this pattern have now
been well documented. Recent phylogenetic. stud-
ies have confirmed the close affinity of the inter-
continental disjuncts in many groups (reviewed in
Wen. 1999), but most presumed species pairs stud-
ied so far between the two continents do not show
a direct sister-species relationship. Morphological
stasis (i.e... the lack of significant morphological
change in a species over long periods of time) has
been suggested as a common phenomenon among
these disjuncts (Parks & Wendel, 1990; Qiu et al..
1995a, 1995b) and may at least partially explain
distribution of
Arrows indicate isolated areas of distebütiong
Distribution of Osmorhiza sect. Osmorhiza, showing Asian
f O. berteroi showing American antitropical, and eastern—
To test
the reported. findings (Wen, 1999, 2001).
this hypothesis, Wen (1998) suggested the exami-
nation of eastern. Asian—eastern North American
disjunct taxa within the context of biogeographic
studies covering the Northern Hemisphere or even
at a global scale. In-depth comparative studies of
disjunct groups that also have apparent close rel-
atives in western North America may provide in-
sight into the relative morphological, molecular,
physiological, and cladogenetic rates of evolution
of taxa in all three areas.
\ntitropical disjunctions between western North
America and southern South America were recog-
nized as early as 1880 by Gray and Hooker, and
have been discussed by many authors since (e.g..
Engler, 1882; DuReitz, 1940; Campbell, 1944:
Constance, 1963; Raven, 1963; Cruden, 1966;
Moore, 1972; Thorne, 1972, 1993; Cox, 1990; Pe-
terson & Morrone, 1997; Peterson & Ortiz-Diaz,
998). Constance (1963) pointed out that the spe-
cies exhibiting this pattern comprise a highly un-
—
representative sample of the floras in the two dis-
that
self-compatible and often autogamous. Based on
junet areas, and many of them are
416
Annals of the
Missouri Botanical Garden
these and other considerations, Raven (1963) con-
cluded that the most plausible explanation for the
observed antitropical disjunctions is relatively re-
cent long-distance dispersal, although Cruden
(1966) presented arguments in favor of “mountain
hopping” that involves shorter distance dispersal
along the western cordillera.
Fernald (1924, 1925,
attention to disjunctions between western North
1935) was the first to draw
America, the Great Lakes region, and northeastern
North America, which he thought involved primar-
ily arctic and western taxa that were able to survive
during the Pleistocene in unglaciated areas (“nun-
taks”) around Lake Superior, on the Gaspé Pen-
but
Й and in Newfoundland and Labrador,
3
could not uy their range following the glacia-
tions because of their antiquity. However, all of Fer-
nald's nunataks were in fact glaciated, and the no-
tion of senescent species is now widely rejected
(Wood, 1972
was proposed by Schofield (1969), who regarded
). A modified version of this hypothesis
the eastern populations as remnants of a previously
widespread flora whose members were able to sur-
vive only in some areas (probably south of the gla-
cial boundaries, but perhaps also in nunataks) and
then move to their present sites following the Pleis-
tocene, while being eliminated from the refugia.
Lowry and Jones (1984) viewed this as a plausible
explanation for the western North American—east-
ern North American disjunctions in О, berteroi and
O. depauperata, and further suggested that taxa oc-
curring in the Great Lakes area and the northeast
are most likely now restricted to sites with less
competition from eastern boreal taxa and where cli-
matic conditions (especially snowfall and moisture
availability in spring) are similar to those in west-
ern parts of the continent.
Another fascinating aspect of Osmorhiza is that
the type of seed dormancy differs among the spe-
cies, and in particular between the putative close
relatives in eastern Asia and eastern North Amer-
ica. Very little comparative research has been done
on physiological traits of species that exhibit this
classical disjunct pattern (Terui & Okagami, 1993;
Wen et al., 1996),
ducted within a phylogenetic framework (Wen et
with only one investigation con-
al., 1996). Studies on seeds of five Osmorhiza spe-
cies found that they have morphophysiological dor-
mancy (MPD) (Baskin & Baskin, 1984, 1991: Bas-
1995; Walck et al., 2002). Seeds with
MPD have embryos that are very small relative to
kin et al.,
the size of the seed, and the embryo must grow to
the full length of the seed to germinate. In addition
to being small, the embryos also have physiological
dormancy that must be broken before seeds can
1998). Seeds of О.
claytonii and О. longistylis have nondeep complex
germinate (Baskin & Baskin,
MPD, i.e., they require warm followed by cold strat-
ification to germinate, and gibberellic acid (GA,)
substitutes for warm stratification (Baskin & Bas-
kin, 1984, 1991). In contrast, seeds of O. aristata
and those from western North American popula-
tions of О, berteroi and O. occidentalis have deep
complex MPD, requiring only cold stratification to
overcome dormancy, and GA, does not substitute
for stratification (Baskin et al., 1995; Walck et al.,
2002). Thus, mapping the types of MPD on a phy-
logenetic tree can provide an opportunity to eval-
uate the evolution and adaptive significance of this
physiological trait.
‘onstance and Shan (1948) proposed an infra-
generic classification for Osmorhiza that was largely
adopted by Lowry and Jones (1984) with a few
modifications (Table 1). Several hypotheses of re-
lationships are implied in this classification. Two
subgenera are recognized, one of which (subg. Gly-
cosma) comprises a single western North American
species, O. occidentalis, which would thus repre-
sent the sister species to a clade that includes the
remaining nine species. Within the typical subge-
nus, three sections are defined (each with three
species), which are likewise assumed to comprise
monophyletic groups. This classification thus offers
1 framework against which to test alternative hy-
potheses of relationships using molecular sequence
data that were not available previously.
The objectives of this study are to: (1) recon-
struct the phylogeny of Osmorhiza using sequence:
of the internal transcribed spacer (ITS) and 5.85
regions of nuclear ribosomal DNA; (2) examine the
biogeographic diversification within the group: (3)
re-evaluate the earlier hypotheses regarding the
possible origins of the observed disjunct distribu-
tional patterns; and (4) examine the evolution of
seed dormancy in a phylogenetic framework for dis-
junct and
type(s) of MPD are plesiomorphic versus derived.
taxa of Osmorhiza determine which
Sequences of ITS and 5.85 regions were employed
because they have been shown to be appropriate to
assess evolutionary relationships within other
groups exhibiting disjunct distributions between
Asia and North America (Fritsch, 1996; Lee et al.,
1996; Wen & Zimmer, 1996; Wen et al., 1998) as
well as interspecific relationships in Apiaceae
1998)
other north temperate. plants (e.g.. Xiang et a
1998; Lee & Wen, 2001).
(Downie et al., and evolution within many
MATERIALS AND METHODS
Forty-eight populations representing all ten spe-
cies of Osmorhiza and three outgroup taxa were
Volume 89, Number 3
2002
Wen et al.
Osmorhiza (Apiaceae)
Table
(1984), as modified from Constance and Shan (
Taxa of Osmorhiza Raf. and their distributions following the classification scheme of Lowry and Jones
8)
Classification
Taxon
Distribution
Subgenus Glycosma таш Drude
Subgenus Osmorhiz
O. brachypoda Torr.
O. glabrata Phil.
О
mexicana Griseb.
Section Mexicanae п ИНВ
& Shan ex Lowry & A. С. Jones
subsp. bipatriata (Constance & Shan)
wry € A. G. Jones
subsp. mexicana
Section Nudae Constance & Shan 0. berteroi DC.
& G. Jones
ex Lowry ¢
O. depauperata Phil.
O. purpurea (J. M Coult. & Rose)
Suksd.
Section Osmorhiza
). claytonii (Michx.) €
o longistylis (Vorr.) DC.
O. occidentalis (Nutt.) Torr.
s aristata (Thunb.) Rydb.
W North America
California, Nevada, and Arizona
central Andes
SW Texas and N Mexico
N Mexico to N Argentina
W North America, Great Lakes area,
NE North America, and South
America
W North America, Great Lakes area,
NE North America, and South
America
NW North America
te — Asia
. B. Clarke EN
E Lu America
rth America
sampled in this study (Table 2). Populations of
widespread disjunct taxa such as O. berteroi and O.
depauperata were examined from throughout much
of their distributional ranges. Anthriscus caucalis M.
Bieb.,
Engstrand.
chosen as the outgroups because of their close re-
1984:
Geocaryum macrocarpum (Boiss. & Spruner)
and Myrrhis odorata (L.) Scop. were
lationship to — za (Lowry & Jones,
Downie et al., 200
Total DNA was — with the CTAB method
of Doyle and Doyle (1987) and purified over EsCl/
DNA amplifications
were performed in 100-j.L reactions following Wen
and Zimmer (1996) using the primers С26А and
Nncel8810 (see Wen «€ Zimmer for sequences of
The entire ITS and 5.85 regions were se-
ethidium bromide gradients.
primers).
quenced manually from both directions following
Wen et al. (1998) using four primers: (5.85. C264.
ITS4, and NI8LI8 (see Wen € Zimmer for se-
quences of primers).
The DNA sequences obtained were assembled.
and the boundaries between the coding and spacer
regions were determined by comparing them with
the sequences of Daucus carota L. (Yokota et al.,
1989). The sequences were then exported to
PAUP* (vers. 4.0, Swofford, 1999).
substitutions, allowing
Most mutations
were base thus manual
alignment. All the sequences have been deposited
at GenBank (see
r
Table 2 for accession numbers).
Phylogenetic analyses were performed with
PAUP* using maximum parsimony (Swofford et al..
maximum likelihood (Felsenstein, 1981).
1987) methods.
Parsimony analysis was performed using a branch-
and-bound search with MULPARS and furthest ad-
dition sequence options. The amount of support for
1996),
and neighbor-joining (Saitou & Nei,
monophyletic groups revealed in the maximally
parsimonious tree(s) (MPTs) was examined with
1000 bootstrap replicates (Felsenstein, 1985) with
random addition and heuristic search options. The
maximum likelihood analysis was performed with
the input order of sequences randomized and the
transition/transversion ratio set at 1.42 based on the
observed frequencies in the MPTs of the parsimony
А neighbor-joining tree was constructed
distance. (Kimura.
analvsis.
using Kimura. two-parameter
1980)
Relative rate tests were performed with the meth-
od of Wu and Li (1985) to detect any rate asym-
metries of the ITS and 5.85 regions among taxa in
Osmorhiza. The proportions of site differences were
estimated using the Kimura two-parameter distance
(Kimura, 1980).
The optimal area cladogram was constructed
from taxon cladograms using the optimality method
in COMPONENT (vers. 2.0, Page. 1993). The fol-
lowing options were used: nearest-neighbor inter-
changes and minimizing the number of leaves add-
endemism were defined for
ed. Four areas of
Osmorhiza based on the distribution of taxa and
previous biogeographic studies of the North Tem-
1970: 1981:
perate zone (e.g.. Wood, Patterson.
418
Annals of the
Missouri Botanical Garden
V86ESVAV
C86ESPAV
C86CSPAV
8LOESPAV
I86€£€T.4 V
086€S€T AV
6L6€ST AV
LLOESPAV
9LOESVAV
€L6ESVAV
CSOESPAV
UPA] оләпұ “ODIXIJA
TA ONW *oorxo]y
ору “ODIXIA
`о7) 3pga&e 4 *&xyonjuaN "sn
`07) e3eieg “URBINA ус
°09) PANPIEA, "еце" цроу "V 5 :
°0) SATIN eming "Vig
OpeloW [4 Nd *o2enueq Р 31U21 4 әр; :
вое, әр “4014 ‘OYO
&a][PA ешүеәт “opto
07) snowy *oaqon() peue
`о7) 1әшилР^] *opea1opo?) “VS
03 SPJUOYRIOG “PIUIBATA 1524 “VS
°07) INAL J "AXonquaN "VS
'07) лоше *pue[&re|y “YSA
o7 e3e1eg “URSIYOI “VSN
707) шер "erueA[4suuag “y Sr]
707) ээри “AYonjusy “VS
оу зип wuuepnm y) VST
‘09 uone *uo224() ус
°03 SPUR YORI) *uo3a4() “y SN
*07) 9dse«) *ooqon() “epeue”)
'07) snowy *23qon() *epeue)
o9 om] *ue3rjory “W'S
‘o9 mng ‘вішоуео "vis n
‘ory — "PIUdOJI[e7) “WIS
‘Jug Uouov dar
‘AOIg Buetfey7 ‘euyy
(4) 2816 poumy f W X 1 тот а d
(SO) 9/21 Funryy 4
(SD) (86-1A-01) "ws SUDIQU Hs
(4) £99 UDADPIH "NES Y PPM TT
(4) с/р vay T
(V) 198 vay f
(ү) 228 vay f
(OW) 61 FZ 1” 19 12021 "S
(SO) 6131 210421 “Wx
(SO) 8111 24042 "IW
(4) crzr voy f
(4) orcg uM Tx
(A) 29€ охор ^N "S Y PPM 71 f
(4) cog UDÁDPIY N 7S X PPA 7T f
(SD) ESE РИО А
(OW) 0667 7” 12 1] мот d d
(OW) 6f6F JP 1? [] тот d d
(SO) 8t&6 PIPNSO ‘H A
(S9) 014€ ә] `q X «omoupf Y 7]
(4) 2211 чәзппоу 7») Y vzavg ^w
(4) GSES] `P 12 Assanig [
(4) 929 Y9S1NOY 7) ӯ гәр "y
(A) €29 asoy ©) Y эгэр "Ws
(4) us ung M ^g Y 977 HT
(4) Yes Dang2uoq ^N
(4) "ws NOYZ “Sy
Dnjmuuqpdiq ‘dsqns рирлхәш cf,
: | i 0
DUDIIXIUL 'dsqns Dunpoixout `0
siasiduo] °0
070190]8 °0
pımıadnndap ‘0
1101012 `0
vpodíy. 2010) ( )
10191199 '()
DIDISUID `0
U0ISSIJIB xueg Ud‘)
JIINOS
19YINOA
ПОХР],
“(€ 3[qep 99S) S]S9] Ә]Р1 PAN POI əy} ur posn әләм whe чим po»xreur suornoo[[o?) 'pe[duies dnoi3jno 34] pue sardads 021410ш5() jo SUOISSIIOY A 3[4B.L
Volume 89, Number 3
002
Wen et al. 419
Osmorhiza (Apiaceae)
Continued.
Table 2.
GenBank accession
Source
Voucher
Taxon
AF453985
AF453986
orado, Garfield Co.
U.S.A., Co
*B. Painter et al. 18 (CS)
J. Wen 3860 (F)
O. occidentalis
— (MN mM b-
oO © е ¢
DD DQO
uU» min ш
E ы ш a
333
a з
„оо C .
2255
ы Itc
o >
=e Vo
AO. € x
Laon. # oos
eod dd ue
т” ЖЫ = ue s
BEES
bef a
© зы d =
с "a “a 0
осоо
4j uiui d
unt
[db Mi
= =
™ N
QN Mm
+s
DO
€ دہ 2
ES
> ©
=
NOUS
AS
‘Sy ой +
bip ые Ше
os S
Rk 8 sa
6 = = 5
о S S &
س о A б
— AA
A д
= ITA
Seen
U
S.A., Nevada, Lander Co.
AF453989
К. К. Halse 5561 (CS)
U.S.A.. Oregon, Benton Co.
U.S.A
R. R. Halse 5565 (CS)
AF453994
AF453995
Oregon, Tillamook Co.
U.S.A., Oregon, Linn Co.
+
P. P. Lowry 11 4963 (MO)
P. P. Lowry 11 5069 (MO)
O. purpurea
AF453996
U.S.A., Oregon, Linn Co.
Downie et al. (1998)
070 (MO)
5
P. P. Lowry Il
Anthriscus caucalis
3605. AF073606
901
AFO
AFO
(2000)
Downie et al.
Geocaryum macrocarpum
ad
Valiejo-Roman et al. (1998)
Myrrhis odorata
Hoey & Parks, 1991), work on Osmorhiza (Lowry
& Jones,
analysis. These areas are western North America,
1984), and results of this phylogenetic
eastern North America, central Andes, and Asia.
South America was not chosen as one of the areas
of endemism because it is linked to western North
America via disjunction and shares the same spe-
cies, suggesting that it was recently colonized (also
see Discussion).
RESULTS
CHARACTERISTICS OF ITS SEQUENCES
The combined length of the ITS], 5.85, and 1782
regions in Osmorhiza species 15 005 bases, with an
ITS] of 217-220 bases, a 5.85 of 164 bases, and
an ITS2 of 224-225 bases. Two insertions in ITS]
were required to align the sequences within the ge-
nus. One occurs in the three accessions of Asian
О. aristata (1 bp), and the other is a 3-bp insertion
1 O. mexicana subsp. bipatriata. Two additional
insertions are in ITS2, each 1-bp in length. The
alignment of sequences of Osmorhiza with those of
the outgroup species required five additional 1-bp
and one 2-bp indels, and sequences of Osmorhiza
and the outgroups could be aligned manually with-
out difficulty. The sequence divergence of Osmor-
hiza taxa and the outgroups ranges from 4.679 (be-
tween Myrrhis odorata and О. berteroi) to 14.05%
(between Anthriscus caucalis and O. mexicana
subsp. mexicana).
PHYLOGENETIC ANALYSES
Of the 693 aligned positions, 116 sites were var-
iable: of these, 62 were phylogenetically informa-
tive. Treating gaps as new characters, the parsi-
mony analysis generated a single most
parsimonious tree (MPT) with a total length of 157
steps, a consistency index (CI) of 0.860 (0.763 ex-
cluding uninformative characters), a retention in-
dex (RI) of 0.912, and a rescaled:consistency index
(RC) of 0.785 (Fig. 2). Treating gaps as missing
data, the parsimony analysis produced a single
MPT an identical topology to Figure 2 with a
141 steps, a Cl 0.851 (0.756 when
se abel. characters are excluded), an. RI
0.974, and an RC = 0.916. The tree from the
weighted parsimony analysis (weighting transver-
sions over transitions 1.42 times) also had an iden-
tical topology to the MPT shown in Figure 2. Sev-
length «
eral relationships are suggested by the parsimony
analyses: (1) the three populations of O. aristata
from Asia form a monophyletic group: (2) O. clay-
Annals of the
Missouri Botanical Garden
10
100
18
—
1
«50
C2
of of of of o
eo
©
— —
<50
50
Nondeep Complex
MPD Clade
alis
eocaryum macrocarpum
С Myrrhis odorata
aristata China
aristata Japan
aristata Korea
rteroi675
5 [South
(c
berteroi15556|
berteroi4726 fe
berteroi4732 Quebec
berteroi4949 Oregon
berteroi4950 Oregon
[
l
l
[
E Asian
Clade
oo
об
22
9
o
N
berteroi5710 California
berteroi9248 California
berteroi4707 Michigan
brachypoda9253 California
depauperata1118 S America
W N American Clade
d s9493 California
dentalis4633 Nevad
occidentalis5155 Nevada
dentalis5561 —
dentalis5565 Ore
occidentalis9532 California
purpurea4963 Oregon
purpurea5069 Oregon
purpurea5070 Oregon
past plea —
mexic
mex кө " bipatria ta
ytonii565 Kentucky
[567 West Virginia
laryland
Mexico
i812 V
11069 Kentucky
¡1089 Pennsylvania
¡4716 Michigan
longistylis5 64 P Kentucky
ngistylis 827 Virginia
longistylis861 N Carolina
longistylis4715 Michigan
1219
E М American Clade
Central
a2419 Andean Clade
Outgroups
Eurasia
Figure 2. The single most i tree of Osmorhiza with a — length of 157 steps, a СЇ of 0.860 (0.763
excluding uninformative characters), an RI of 0.912,
above lines
names are accession numbers as in
and ar
1 RC of 0.785, treating gaps as new characters. Numbers
are br ranch lengths and those below are bootstrap v: ues 's in 1000 replicates. Numbers following the taxon
Table 2 to help identify the source of study material. The solid black bar at the
node of the О. — O. longistylis clade indicates the most parsimonious explanation of the evolution of the nondeep
complex morphophysiologic: al dormaney (MPI
tonii and O, longistylis from eastern North America
American clade); (3) the
multiple populations examined of the widespread
form a clade (the
O. depauperata form a monophyletic group; (4) O.
depauperata and O. occidentalis form a clade: (5
O. brachypoda and O. purpurea are allied with the
O. depauperata—O. occidentalis clade; (6) popula-
tions of the widespread O. berteroi form a mono-
phyletic group; (7) several species (O. berteroi, О.
brachypoda, O. depauperata, O. mexicana, O. oe-
cidentalis, and O. purpurea) form a largely western
North
pauperata extending to eastern North America and
American clade with O. berteroi and O. de-
South America, and O. mexicana ranging from Tex-
as to South America (the W N American clade);
and (8) О, glabrata from the central Andes forms
Volume 89, Number 3
2002
Wen et al.
Osmorhiza (Apiaceae)
Table 3.
at the 5% level.)
Relative rate tests to detect rate asymmetry. Myrrhis odorata was used as the reference taxon. (* Significant
Species 1 Species 2 K; K K,-K,, + SE
O. aristata O. berteroi 0.06582 0.05165 0.01417 + 0.00611*
О. aristata O. claytonii 0.06582 0.06015 0.00567 + 0.00619
O. aristata O. depauperata 0.00582 0.06759 —0.00177 + 0.00701
O. aristata O. glabrata 0.00582 0.05347 0.01235 + 0.00545*
O. aristata О. mexicana 0.060582 0.0644 0.00140 + 0.00669
O. aristata O. occidentalis 0.06582 0.06993 —0.00411 + 0.00752
O. berteroi O. claytonii 0.05165 0.06015 — 0.00850 + 0.00499
O. berteroi O. depauperata 0.05165 0.06759 —0.01594. + 0.00450*
O. berteroi O. glabrata 0.05165 0.05347 —0.00182 + 0.00397
O. berteroi O. mexicana 0.05165 0.06442 —0.01277 + 0.00398*
O. berteroi O. occidentalis 0.05165 0.06993 —0.01828 + 0.00493*
O. claytonti O. depauperata 0.06015 0.06759 —0.00744. + 0.00619
O. claytonii O. glabrata 0.06015 0.05347 0.00668 + 0.00415
O. claytonii O. mexicana 0.06015 0.06442 —0.00427 + 0.00574
O. claytonii O. occidentalis 0.06015 0.06993 —0.00978 + 0.00641
O. depauperata O. glabrata 0.06759 0.05347 0.01412 + 0.00520*
O. depauperata O. mexicana 0.00759 0.00442 0.00317 + 0.00481
О. depauperata O. occidentalis 0.06759 0.06993 —0.00234. + 0.00347
O. glabrata O. mexicana 0.05347 0.06442 —0.01095 + 0.00477*
O. glabrata O. occidentalis 0.052347 0.06995 —0.01646 + 0.00550*
O. mexicana O. occidentalis 0.06442 0.06993 —0.00551 + 0.00568
a trichotomy with the E N American clade and the
W N American clade.
The maximum likelihood tree (MLET, not shown,
with a log likelihood of = 1512.26) has an identica
topology to the MPT. The neighbor-joining tree
(NJT) is similar to the MPT and the МЇЛ but differs
in that (1
) O. glabrata is placed sister to the O,
claytoniti-O. longistylis clade in the NJT: and (2)
the O. claytonii-O. longistylis—O. glabrata clade is
sister to the Asian О. aristata. The new clades are
weakly supported with bootstrap values less than
50%
SEQUENCE DIVERGENCE
Treating gaps as missing data, the Kimura two-
parameter distance among species of Osmorhiza
was estimated to be 0.209-3.784%.
gence values are comparable to those in
These diver-
lralia
sect. Aralia, which shows a similar disjunct pattern
of distribution, with representatives in eastern Asia
and in eastern and western North America (Wen et
c
al., 1998).
O. occidentalis from western North America and €
The highest divergence occurs between
aristata from eastern China, and the lowest —
O. claytonii and O. longistylis from eastern North
America, which are sympatric throughout much of
their ranges. Overall, divergence values between O.
aristata and other members of the genus are com-
paratively high, ranging from 1.627% (with O. clay-
tonii) to 3.784% (with O. occidentalis). Among the
New World species, sequence divergence of O. gla-
brata from the central Andes ranges from 0.601%
(with O. longistylis) to 2.552% (with O. occidental-
is). The western North American O. occidentalis has
a relatively high level of sequence divergence,
ranging from 0.319% (with O.
2.552% (with O. glabrata). Infraspecific variation
—
depauperata) to
was detected in six species for which multiple sam-
ples were available, O. aristata (O-0.302%), O. ber-
terol (O-O.473%), O. claytonit (O-0.517%), O. mex-
icana (0-1.415%). O. occidentalis (0-1.013%), and
O. purpurea (0-0.148%).,
lations sampled of O. depauperata and O. longis-
whereas multiple popu-
tylis showed no variation in their ITS profiles, de-
fact thal
distributions. The absence or relatively low level of
spite the these species have wide
sequence variation within these taxa supports the
species level cireumscriptions proposed by Lowry
and Jones (1984). On the other hand, infraspecific
occidentalis was 1.013% be-
tween populations from the Rocky Mountains and
variation within O,
those from California.
RELATIVE RATE TESTS
Twenty-one relative rate tests (Wu & Li, 1985)
were conducted to detect rate asymmetry (Table 3).
Myrrhis odorata was used as the reference taxon.
Rate differences of most pairs of species were not
Annals of the
Missouri Botanical Garden
“7?
0 — YY МА
3B 75'E са
5 $
Figure 3. Biogeography of Osmorhiza.—A. Area cladogram of Osmorhiza based on the nuclear ribosomal ITS
phylogeny. —B. Model of diversification in Osmorhiza.
statistically significant at the 5% level, whereas
eight species pairs showed significant differences.
These tests suggest that O. berteroi had a slower
rate of nucleotide substitutions in the ITS regions
than O. aristata, O. depauperata, O. mexicana, and
O. occidentalis. They also show that O. glabrata
had a slower rate of nucleotide substitution than O,
aristata, O. depauperata, O. mexicana, and O. oc-
cidentalis. The molecular clock hypothesis for the
ITS sequences in Osmorhiza was therefore rejected.
AREA CLADOGRAM CONSTRUCTION
The area cladogram (Fig. 3A) was constructed
with the MPT and MLT using the optimality method
in COMPONENT (vers. 2.0, Page, 1993). Among
the four areas of endemism, Asia is basal and east-
ern North America, western North America, and
central Andes form a trichotomy.
DISCUSSION
PHYLOGENY AND PATTERNS OF DIFFERENTIATION
'arsimony and maximum likelihood analyses
support the basal position of the Asian O. aristata
within the genus and the monophyly of the New
World taxa; these relationships were previously
suggested in Downie et al. (2000) in a broader anal-
ysis of the tribe Scandiceae of Apiaceae. Within
the New World species, three major subclades are
suggested: (1) the eastern North American O. clay-
tonii and O. longistylis; (2) a largely western North
American clade comprising O. berteroi, O. brachy-
poda, O. depauperata, O. mexicana, O. occidentalis,
and O. purpurea; and (3) the central Andean en-
demic O. glabrata. The neighbor-joining tree, how-
ever, groups the central Andean endemic O. gla-
brata with the eastern North American O. claytonii
and O. longistylis. The O. claytonii-O. longistylis—
O. glabrata group is then sister to O. aristata from
Asia. As in the MPT and the MLT, the NJT also
suggests a largely western North American group
comprising O. berteroi, O. brachypoda, O. depau-
perata, O. mexicana, O. occidentalis, and O. pur-
purea.
The ITS phylogeny conflicts in several aspects
with the phylogenetic hypothesis implied in the in-
frageneric classification of Lowry and Jones (1984)
based on morphology (cf. Table 1). Osmorhiza oc-
cidentalis is a morphologically distinct species and
was thus originally described in the monotypic ge-
nus Glycosma by Nuttall (in Torrey & Gray, 1840).
Recent treatments have included this species with-
in Osmorhiza, but placed it in its own section (Con-
stance & Shan, 1948) or subgenus (Lowry & Jones,
1984) because of the numerous features that dis-
tinguish it from other members of the genus, in-
cluding glabrous fruits lacking a caudate append-
age. numerous staminate umbellules and flowers,
bipinnate leaves, and yellow to greenish yellow
flowers. The inclusion of O. occidentalis within a
well-supported western North. American clade
the ITS phylogeny suggests, however, that its dis-
Volume 89, Number 3
2002
Wen et al. 423
Osmorhiza (Apiaceae)
tinctive morphological characters are most likely
attributable to rapid evolution within the species,
resulting in many autapomorphies.
Another discrepancy between the current infra-
generic classification and the ITS phylogeny in-
volves the three species currently included in Os-
morhiza sect. Osmorhiza: the Asian O. aristata and
the eastern North American O. claytonii and O.
longistylis. These taxa form a morphologically co-
herent group and were even treated as a single spe-
cies by some authors (e.g.. Gray, 1859; Clarke.
1879; Kuntze, 1891; Boivin, 1968). They share sev-
eral diagnostic features, including ап involucre
composed of (12)2-3(-5) conspicuous, foliaceous
bractlets, and styles (including the high-conic sty-
lopodium) that are 1-3.6 mm long. However. based
on ITS data, O. aristata is not only basal within the
genus, but also shows a high level of sequence di-
vergence from its congeners (1.627—3.784%), sug-
gesting its early divergence within the group. The
relative antiquity of O. aristata and the compara-
tively high level of morphological similarity. be-
tween it and the eastern North American species
are consistent with the hypothesis of morphological
stasis among eastern Asian—eastern North Ameri-
can disjuncts, as proposed by Parks and Wendel
(1990).
are sister to the western North American clade and
Taxa of the eastern North American clade
not to O. aristata. The morphological similarity be-
tween 0. aristata, O. claytonit, and O. longistylis
could, however, be explained by their having ex-
perienced morphological stasis while concurrently
the largely western North American clade diversi-
fied more rapidly.
Morphological stasis was first proposed by Parks
and Wendel (1990) as a possible explanation for
the presence of morphological similarities among
eastern Asian—eastern North American disjunct
taxa that are not closely related or do not represent
sister groups. They found that two morphologically
similar disjunct species of Liriodendron (L. chinense
(Hemsley) Sargent and L. tulipifera L.) show a high
level of allozyme and cpDNA divergence. Molecu-
lar and fossil data suggest the divergence time of
mid-
—
the two species to be 10-16 million years ago
dle Miocene). Morphological stasis has also been
suggested to explain similarities between the Asian
and North American species of Aralia sect. Dimor-
phanthus (Wen, 2000). Liquidambar (Hoey
Parks, 1991; Shi et al., 1998), and Magnolia sect.
Rytidospermum (Qiu et al., 1995a, 1995b)
A third conflict between the current infrageneric
2
classification of Osmorhiza and the ITS data in-
volves the central Andean endemic 0. glabrata.
The relationships and taxonomic placement of this
species have been somewhat ambiguous. Constance
and Shan (1948) included О. glabrata in their sec-
tion “Glycosmae” along with O. occidentalis, largely
based on apparent similarities in their fruits. By
contrast, Lowry and Jones (1984) placed the An-
dean endemic in section Mexicanae, which also
comprises O. brachypoda and O. mexicana. Shared
characters among these species include short styles
(0.5—1.2 mm long), a low-conic stylopodium, a con-
spicuous involucre of 1—6 bractlets, and an absent
or poorly developed involucel subtending the um-
bellules. By contrast, the maximum parsimony and
the maximum likelihood analyses suggest that О.
glabrata forms a trichotomy with the eastern North
American clade (О. claytonii and O. longistylis) and
The
neighbor-joining tree, however, places O. glabrata
the largely western North American. clade.
eastern North American O. claytonti—
The ITS phylogeny (Fig. 2)
thus suggests the relative antiquity of O. glabrata
sister to the
O. longistylis group.
among the New World species. It seems most plau-
sible that O. glabrata was derived from a North
American ancestor, perhaps shared with O. clay-
tonii and O. longistylis (as in the NJT) or with taxa
of the eastern North American clade and the west-
ern North American clade (Fig. 2). The present nar-
row distribution of O. glabrata in the central Andes
may be due to local survival following the Pleisto-
cene glaciations, which have been considered im-
portant in the evolution of the Andean flora (Vuil-
1971; 1975, 1983;
1982). Additional data are, however,
leumier, Simpson, Prance,
needed to as-
certain the exact phylogenetic position of O. gla-
rata.
INFRASPECIFIC VARIATION
Differentiation within O. occidentalis. Osmorhiza
occidentalis showed a relatively high level of infra-
specifie sequence divergence, with a Kimura two-
parameter distance of 1.013% between populations
from the Rocky Mountains and California. Morpho-
logically, O. occidentalis is somewhat variable, and
several segregate species have been recognized in
the past based primarily on differences in inflores-
cence structure, leaf pubescence, and fruit size (see
1984).
were regarded as minor by Lowry and Jones (1984),
These differences, however,
owry & Jones,
who provided data showing that a broadly defined
O. occidentalis comprises a morphologically coher-
ent group despite the presence of a few rather atyp-
ical collections (some of which served as the basis
for the segregate species). Because O. occidentalis
is so distinct from its congeners, it may be that
infraspecific morphological variation has been over-
424
Annals of the
Missouri Botanical Garden
looked by previous workers. Additional studies
should be undertaken to re-assess variability within
O. occidentalis and to evaluate whether the molec-
ular divergence is correlated with morphology and/
or geographic distribution.
Differentiation between O. mexicana subsp. mex-
icana and subsp. bipatriata. The two subspecies of
O. mexicana show a relatively high level of ITS
sequence divergence (1.415%). Moreover, they did
not form a monophyle tic group in the phylogenetic
analysis (Fig. 2). Initially, Constance and Shan
1948) recognized these taxa as distinct species,
but Lowry and Jones (1984) treated them as sub-
species because of the presence of occasional mor-
n
„ч.
phological intermediates at several localities
northern Mexico, including one area (Cerro Poposí,
Nuevo León) where they co-occur with typical ma-
terial of both subspecies. The biphyly of O. mexi-
cana, as currently circumscribed, suggests the need
to reexamine the status of subspecies bipatriata and
the possible causes of the morphological interme-
diates. Although no interspecific hybridization has
been reported for Osmorhiza, occasional morpho-
logical intermediates have been observed in sym-
patric populations of O. berteroi and O. occidentalis
in western Oregon (R. Halse, pers. comm.).
BIOGEOGRAPHY
Diversification in major geographic areas. The
ITS phylogeny (Fig. 2) shows that rates of clado-
genesis vary among the four major areas occupied
by Osmorhiza: eastern Asia, eastern North America,
western North America, and the central Andes.
Cladogenesis clearly appears to have been more
rapid in the New World than in Asia, and the west-
ern North American clade shows a particularly high
level of species diversity (Fig. 3B). It is unusual
that Asia is species-depauperate considering that
most other Asian—North American disjunct genera
show a higher species diversity there (Tiffney,
1985; Wen, 1999). The number of Osmorhiza spe-
cies occurring in each area may in part be the re-
sult of differential rates of extinction, especially
during the Quaternary glaciations. However, it is
generally accepted that North America was more
severely affected by glaciation than eastern Asia
due to its more complex topography and the north—
south rather than primarily east-west orientation of
its mountain ranges (Axelrod et al., 1998; Wen,
¢ .
Western North America is the center of species
diversity in Osmorhiza, with six of the ten species
occurring in this region. The monophyly of the
western North American clade is well supported in
all three analyses presented here. Morphologically,
however, the clade is rather diverse, with members
from both subgenera and two of the three sections
recognized by Lowry and Jones (1984). The ITS
phylogeny suggests two successive diversifications
within Osmorhiza in western North America. First,
the common ancestor of the western North Ameri-
can clade may have diversified into O. berteroi, O.
mexicana, and the ancestor of the O. brachypoda—
O. depauperata—O. occidentalis-O. purpurea sub-
clade. The latter subclade then appears to have dif-
ferentiated further. This diversification among
western North American Osmorhiza may have been
highly influenced by the availability of a range of
habitats associated with the uplifting of the Rocky
Mountains and the western cordillera during the
Tertiary (Barbour & Christensen, 1993; Graham,
1993).
The Andean endemic O. glabrata most likely
represents an isolated relict. Its phylogenetic po-
sition is not well-resolved in the ITS phylogeny. It
may have diverged rapidly early in the evolutionary
history of the North American Osmorhiza, persist-
ing in a relatively small portion of the central An-
des. The fruits of O. glabrata are glabrous to
sparsely hispid, which may afford them limited op-
portunities for long-distance dispersal by animals.
It should be noted, that O. occidentalis
almost always has completely glabrous fruits but
however,
nevertheless extends throughout a much larger area
of western North America (Lowry & Jones, 1984).
Eastern Asian—eastern North American disjunc-
—
tion. The relative antiquity of the eastern Asian—
eastern North American disjunction in Osmorhiza
is suggested by the phylogenetically basal position
of the Asiatic O. aristata, the monophyly of the
diverse New World species (Fig. 2), and the rela-
tively high ITS sequence divergence between О,
aristata and its congeners. The most commonly ac-
cepted interpretation of the origin of eastern Asian—
eastern. North American disjunctions is that this
pattern reflects an initial widespread distribution of
temperate forest elements in the Northern Hemi-
sphere during the mid Tertiary followed by subse-
quent extirpations in. western North America and
western Europe as a result of late Tertiary and Qua-
ternary climatic cooling (Graham, 1993; Manches-
ter, 1999; Wen, 1999). No fossils of Osmorhiza have
been reported that could help to date the biogeo-
graphic disjunction. Also, the rejection of the mo-
lecular clock hypothesis for the ITS sequences in
Osmorhiza (see Results) makes it questionable to
attempt an indirect estimate of the times of diver-
gence between the eastern. Asian and the North
American members of the genus.
Volume 89, Number 3
2002
Wen et al. 425
Osmorhiza (Apiaceae)
The apparent close relatives of Osmorhiza, the
genera Myrrhis and Geocaryum (Downie et al.,
2000),
are restricted to the Old World, which is
consistent with a hypothesized Old World origin of
Osmorhiza (Downie et al., 2000). The basal position
of the Asian O. aristata and the Old World distri-
bution of Myrrhis and Geocaryum are consistent
with the idea that the ancestor of the North Amer-
ican Osmorhiza migrated from Asia.
American antitropical disjunction. Osmorhiza
berteroi and O. depauperata show a similar pattern
of antitropical disjunction between western North
America and temperate South America. Constance
(1963) suggested that these species may have mi-
grated south in a step-wise manner along the west-
ern American cordillera throughout the Tertiary.
with subsequent elimination of populations from in-
tervening tropical areas. The present-day distribu-
tion of O. mexicana subsp. mexicana, which com-
prises isolated populations that bridge the areas
currently occupied by O. berteroi and O. depauper-
ata, appears to support the idea that the now-dis-
junct species could also have had a more contin-
uous distribution in the past. However. the absence
of ITS sequence divergence between the western
North American and South American populations
of both species suggests a recent origin of the ob-
served M disjunctions. Species of Osmor-
hiza are facultatively autogamous, which would en-
able establishment of a new population from a
single propagule, and their armed fruits with cau-
date appendages and numerous retrorse bristles ap-
pear to be well adapted for epizoochorous dispersal
1984).
(1963), the pattern of disjunction of temperate taxa
North
South America corresponds closely to the migration
(Lowry & Jones, s indicated by Raven
between western America and southern
routes of many bird species, and this probably ac-
counts for the disjunctions seen in О, berteroi and
O. depauperata (Lowry & Jones, 1984). Thus, the
lack of ITS sequence divergence, the monophyly of
both species, and their fruit morphology are all con-
sistent with an origin of the antitropical disjunction
due to relatively recent long-distance dispersal via
birds from western North America to South Amer-
ica. A similar explanation has been used for several
disjunct taxa of grasses (Peterson € Morrone, 1997;
Peterson & Ortíz-Diaz, 1998).
North
The phytogeographic disjunction between
Eastern—western American — disjunc-
tion.
America, the Great Lakes region.
North
cussed repeatedly since it was first reported by Fer-
nald in 1925 (Fernald, 1925, 1935: see also Steb-
1935; 1953; 1969;
western North
and northeastern America has been dis-
bins, Rousseau, Drury,
Schofield, 1969; Morisset, 1971: Voss, 1972: Wood,
1972; Miller & Thompson, 1979).
erally regarded that the disjunct populations of spe-
It is now gen-
cies showing this pattern are remnants of a previ-
ously widespread flora that probably survived south
of the glacial boundaries, but perhaps also in nun-
ataks (cf. Fernald, 1925), and then migrated to their
present sites following the Pleistocene, while being
eliminated from the refugia in the south (Schofield,
1969; Miller & Thompson, 1979). In explaining the
western North American—eastern North American
disjunctions in O. berteroi and 0. depauperata,
Lowry and Jones (1984) also emphasized that taxa
occurring in the Great Lakes area and the northeast
are most likely now restricted to sites with less
competition from eastern boreal taxa and where cli-
matic conditions are similar to those in western
parts of the continent. In the present study, the
western cordilleran populations of O. berteroi
showed an identical ITS profile to those of the Great
Lakes region, while the eastern populations had
The fact that
there is little or no sequence divergence within
just a single nucleotide substitution.
each of these species is consistent with a relatively
recent origin of the disjunct pattern during the gla-
cial or postglacial periods.
PHYSIOLOGICAL ADAPTATION——SEED DORMANCY
Disjunct an ies of herbaceous plants with relict
distributions in. both. eastern. North America and
Asia appear to exhibit stasis in ecological traits
(Ricklefs € Latham, 1992). Similarly, closely re-
ated species groups with members occurring in one
or the other area may also show stasis with regard
lo various traits, including the type of morphophys-
iological dormancy (MPD) they exhibit. This is the
case for seeds of the eastern North American—Asian
disjunct species pairs — diphylla (L.)
Pers.—J. dubia (Maxim.) Ben f. (Ber-
beridaceae) and Panax на L.—P. ginseng
C. A. Meyer (Araliaceae), all of which have deep
simple MPD (cf. Baskin € Baskin, 1998). In Os-
seeds of O. aristata from Asia
morhiza, however,
and 0. berteroi and O. occidentalis from western
North America share the same type of MPD (deep
complex MPD), whereas those of the eastern North
American O, claytonit and O. longistylis exhibit a
different type of dormancy (nondeep complex
MPD). Wake et al. (1983) suggested that changes
in some traits, including physiological ones, might
help an organism to compensate for maintaining a
stable morphology over long periods of time. This
would appear to be the case in members of Os-
morhiza.
426
Annals of the
Missouri Botanical Garden
Although much additional work must be com-
pleted before we have a robust understanding of the
evolution of various types of MPD, some insights
into their evolution may be drawn from the phylog-
eny of Osmorhiza. The plesiomorphic condition in
the genus appears to be deep complex MPD, where-
as nondeep complex MPD is derived. This is fur-
ther supported by studies that show seeds of the
outgroup Myrrhis odorata also have deep complex
MPD (Lhotska, 1977; Deno, 1994). The suggestion
of Baskin et al. (1995) that deep complex MPD may
have been derived from nondeep complex MPD is
not upheld according to the ITS phylogeny of Os-
morhiza.
Seeds of most (and probably all) Osmorhiza spe-
cies are dispersed during summer/autumn, with
germination occurring in spring (Baskin & Baskin,
1984, 1991; Baskin et al., 1995; Walek et al.,
2002). A cold stratification period would thus en-
sure that germination does not occur until the fol-
lowing spring, since dormancy is broken with cold
(winter) stratification. The selective forces respon-
sible for the evolution of the requirement for a
warm stratification pretreatment prior to cold strat-
ification observed in the eastern North American
O. claytonii and O. longistylis are unclear. The most
parsimonious explanation based on the ITS phylog-
eny is that the evolution of this trait occurred in
the common ancestor of these closely related sister
species. It seems reasonable, however, to assume
that a common environmental selection pressure (or
a set of pressures) was responsible for the acqui-
sition of the warm + cold stratification require-
ment. Interestingly, species of Erythronium have a
similar geographic pattern of MPD types, with
seeds from the western North American E. gran-
diflorum Pursh exhibiting deep complex MPD and
those from the eastern North American E. albidum
Nutt. and E. americanum Ker “ж i non-
deep complex MPD (Baskin et al., )5).
Literature Cited
1998. His-
55 in A.-L
Axelrod, D. I., I. Al-Shehbaz € P. H. Raven.
tory of the men flora of China. Pp. 43—
Zhang & S.-G. 1 (editors), Floristic Characteristics
and Diversity of Pd Asian Plants. China Higher Ed-
ucation — Beijing.
Barbour, M. G. N. L. Christensen. 1993. Vegetation.
Pp. 97-1: 3l in F NA Editorial Committee, Flora of North
America North of Mexico, Vol. 1. Oxford Univ.
Ne ^w 'or *
Baskin, С. C. & J. M. Baskin. 1998. Seeds: Ecology, Bio-
ER and pu of Dormancy and Germina-
Aca ademi
E Mant
Press,
tion. © Press, San Diego.
E. al dormancy in seeds of two genera
& J. M. Baskin. 1995. Two types of
(Osmorhiza and e а ап Arcto-Tertiary
distribution pattern. 82: 293-298
Baskin, J. М. & С. С. 84. Cemuinaliun eco-
physiology of the woodland he à —— longistylis
(Umbelliferae). — ч Bot. 37—692.
——— QA ——— к, p come morphophys-
Ismorhiza claytonüi
T
r in
— J. Bot.
3. Flora of the prairie provinces. Phytologia
&
iological
ласеае к,
(Ар 78: 588-593.
Boivin, ig
5 a
eae D. E. & S. A. Spongberg. 1983. Eastern Asian—
astern North American phytogeographic val relation-
ships—A history from the time of Linnae ^i to the twen-
tieth € gr Missouri Bot. Gard. 70: 423—439.
— ll, D. 944. Relations of the te е rate —
с orth me PR America. Proc. Calif. Acad. Sci. 25:
3 —
— 1879. Umbelliferae. Pp. 665-720 in J. D.
eer (editor), Flora of British India, Vol. 2. ewe
Ker
1963 а val relations in the her-
baceous Hora of the Coast of North and South
America: Á symposium. ра — = historical re-
view. zn rt. Rev. n 09-1
—— . H. Sha — n genus Osmorhiza (Um-
be — = © 2 study in pue affinities. Univ. Calif.
deu : 111-15
НА e. L.
—
*
Cox, C. re . New — 'al theories and old biogeo-
phi al pre ms. J. Biogeogr. 17: 117-130.
Cruden, Birds as agents of long-distance dis-
persal or — plant groups of RE tempe rate West-
Evolution 20: 517
Deno, N. C. 1994. Seed Germination: Theory and Prac-
tice, nda a oth ра Published by the author,
State C өн, Pennsylva
Dow 5. RS. panier Katz-Downie & К.
l — 1998. Molecular systematics of Apiaceae sub-
family — ae: Phylogenetic analyses of nuclear ri-
bosomal DNA i Чегин. transcribed spacer and plastid
rpoC | intron sequences. Amer. J. Bot. 85: 563-591
—— ——-, D. S. Katz- — & K. Spalik. 2000. A phylog-
eny of Evidence from nu-
clear — DNA epe — ribed spacer se-
quences. Amer. J. Bot. 87:
Doyle, у 1 & 1 L. Doyle. 1 E is Ai ‘rapid DNA isolation
proc — for — — s of fresh leaf tissue. Phy-
toche m. e -15
Drury, Bi 1969. Plant persistence in the Gulf of
St. Lawrence. Pp. 105-148 in H. Greenidge (ed-
itor), Essays in — deeem and Ecology. Nova Sco-
tia Museum, Halifa
DuReitz, G. E. 1940. а of bipolar plant distribu-
tion. Ac т — орт. Suec. 13: —282.
Engler, ; 382. Versuch einer Entwicklungsgeschichte
dir — ‘пме, insbesondere der Florengebiete seit
der Tertiárperiod. Engelmann, Leipzig.
Felsenstein, J. 1981. Evolutionary trees DNA se-
A maximum likelihood approach. J. Molec.
ern He misphe re.
Apiaceae tribe Scandiceae:
5—
from
quences
Evol. 17: 308-370
1 85 Confide nce limits on phylogenies: ү ар-
proac ‘h — k the bootstrap. Evolution 39: 783-
Fernald, M. L. 1924. Isolation and endemism in —
eastern en rica and their relation to the age-and-area
hypothesis. Amer. J. Bot. 11: 558—573.
—.. 1925. Persistence of plants in unglac iated areas
Arts 15: 239—
T
ә.
of — America. Mem. Amer. Acad.
Volume 89, Number 3
2002
Osmorhiza (Apiaceae)
1935. Critical plants of the Upper Great Lakes
region of Ontario and Michigan. Rhodora 35: 1-16, 47—
120-140. 161-185, 203— — 230-247.
265—283, 298-315, 327-346, 364-386. : )5—403.
Fritsch, P. 1996. Isozyme analysis of interc — dis-
juncts within Styrax (Styracaceae): da rure for the
Madrean- Tethyan hypothesis. Amer. J. Bot. 83: 342
355.
Graham. A. 1972. Floristics and ind deir of Asia
and Eastern North America. Else . Amsterdar
1993. History of the Хаара: Cretaceous
MEI Pp. 57-70 in FNA Editorial
— Flora of North America North of Mexico,
Vo P 7s d Univ. Press, New York.
. Dr. Siebold, Flora Japonica: sectio prima.
— ornatui m usui inservientes; digessit Dr. J. G.
Zuccarini: fasc. ). fol. (A review). Amer. J. Sci. 39:
175-176.
1859. Diagnostic characters of phanerogamous
да ted in Japan by Charles Wright,
Pacific Exploring Expedition, with
observations upon the relationship of the Japanese flora
to that of North America and of other parts of the north-
ern temperate zone. Mem. Amer. Acad. Arts 6: 377-
=
botanist
of the U.S. North I
53.
878. Forest geography = archaeology, a lec-
the val rd University Natural
r. J. Se far ts, ser. 3, 16: 85-94,
ture delivered be " €
History I iety.
183-19
E 1 D. Hooker. 1880. The vegetation. of the
Rocky Mountain region and a Um with that of
other parts " vu world. U.S. Geol. Geogr. Survey Ter-
ritory Bull. 1—77.
oey, M. T. Parka, 199]. [sozyme divergence
between eastern Asian, rth American and Turkish
species ds куы (Hamamelidaceae). Amer. J.
Bot. 947.
Kartesz, 1 T. №. Ghandi. 1993. Osmorhiza berterot (Api-
correct name for mountain sweet-cice y.
1-182.
A simple — for estimating evolu-
aceae): The
Brittonia 15: 18
itution through comparative
. J. Molec. Evol. 16:
tionary rates of base subs
studies of nucleotide sequence
2
1891. Revisio Generum Plantarum. Felix.
Lee, N. S., T. Sang. D. J. Crawford, S. H. Yeau & S.-C.
Ў 1990. Molecular divergence between disjunc t taxa
in eastern Asia and eastern North America. Amer.
Bot. 83: 1373-1378.
Lee. S. € J. Wen. 2001. A phylogenetic analysis of Prunus
and the Amygdaloideae (Rosaceae) based on ITS se-
quences of nuclear ribosomal DNA. Amer. n Bot. 88:
1: —
Lhotska.
M — —
Li, H.-L. 1952. Floristic
Asia and eastern North America.
Soc. 42: 371—429.
Lowry, P. P, H & A. G. Jones. 1984 [1985]. Systematics
of Osmorhiza En (Apiaceae). Ann. Missouri Bot. Gard.
71: 1128-117
Manchester, 5. ^ 1999. Biogeographical relationships of
North American Tertiary oras. Ann. Missouri. Bot.
Gard. 86: 472—5:
Miller, N. G. & G. G. Thompson. 1979. Boreal and west-
. Notes on the ecology of ge rmination in
Phytotax. 12: 209-213.
© relationships between Eastern
Philos.
Folia Geobot.
Trans. Amer.
ern North American plants in the late Pleistocene of
Vermont. J. Pr Arbor. 60: 167—218.
Moore, D. M. 1972. Connections between cool temperate
— тк ‘шаг reference to southern
America. Pp. 115-138 in D. Н. Valentine (editor), Tax-
onomy, мония d and Evolution. Academic Press.
London.
— P. 1971. Endemism in the vascular plants of the
Gulf of St. Naturaliste Canad. 98:
139-149.
Page, Н. D. M. 1993. COMPONENT,
Muset “um ae History), Londor
Parks, 1. & J. F. Wendel. 1990. Molecule divergence
between we and North American species of Liriod-
endron (Magnoliac eae) with — ations of fossil floras.
mer. J. Bot. 77: 1243-12
Patterson, C. 1981. The d ИИ of the North Amer-
ican fish fauna—4A problem of historical biogeography.
Pp. 265-281 in P. E Forey (editor), Chance, Change
and Challenge, Vol. 2. The Evolving Biosphere. British
Museum (Natural — and Cambridge Univ. Press,
London.
Lawre ence region.
Vers. 2.0. British
Peterson, 1997. Allelic
the кык eks al ane t Lycurus setosus (Poaceae:
1-34
Muhle — ae). Madrofio 44: 33 6.
. Morrone. variation in
— ‚ Ortiz-Diaz. 1098. Allelic variation in the
— rop iad disjunct Muhle — ~ yi (Poaceae:
Muhle nbergiinac) Brittonia 50: 381-3
. 1982. Biological ту ‘ation in the Trop-
=
Prance,
ics. Columbia Univ. Press, New York.
Qiu, Y.-L., E — & C. К. Parks.
roplast DNA perinde study of the «
eastern North America disjunct section — rmum
Amer. J. 1 1582-
1995a. A chlo-
astern Аха
e ا (Magnoliaceae).
. R. Parks & M. W. Chase. 1995b. Molecular
divergence in the eastern Asia—eastern North America
disjunct section Rytidospermum of — (Magnoli-
aceae). Amer. J. Bot. 82: 1589-15
Raven, Р. H. 1963. Amphitropical — in the lo-
ras of North and South America. Quart. Rev. Biol. 3
15l-
Ricklefs E. & R. E. Latham. 1992,
corre — of geographic ranges suggests stasis in eco-
logical traits of relict genera of Aero ИЙ perennial
E iue d 139: 1305-1321.
The value of iny as indicator of
unglac dole a areas. Mém. Jard. Bot. Montréal 40; 1-8.
87. The neighbor-joining method:
Intercontinental
=
2
2 =
=
Pp
L
^
ш
=
=
Saitou, N. & M. Nei. 19
A new method v FAR ting phylogenetic trees.
Molec. jus Evol. 25
Schofield, B.
ен EL mds of northwestern
North Veces Bryophytes and vascular plants. Madro-
ño 2( : 5-207
Shi, 5.. H-T. E Cine Y. Chen, L.-H. Qu & J. Wen. 1998.
— пу of the Hamamelidaceae based on the ITS
se d nees of nuclear ribosomal DNA. Biochem. Syst.
Ecol. 25: 55-6
— n . B. 1975. —— changes in $ vun of
the high — al Andes. Paleobiology 1: 273-29
1983. An Жи. алб ШОТ of de hh
Alain Ani Revista Chilena 50: 122.
Stebbins. = L.. Jr. 1935. Some — ations on ni flora
of the Bruce Peninsula, Ontario. Rhodora 37: € 3-7 1.
e . L. 1999. PAUP*: Phylogenetic Analysis Us-
ing Sunderland. Mass
Vers. 4.0. — r,
с is, tts.
428
Annals o
Missouri Botanical Garden
—— Olsen, P. J. bien & D. M. Hillis. 1996.
— aba inference. Pp. 407-514 in D. M. Hillis
Moritz € B. K. Mable (e — Molecular Syste m-
atics, 2nd ed. Sinauer, Sunderland, Massachuse
Terui, K. & N. Okagami. 1993. Temperature кенен оп
seed к germination of east Asian and Tertiary relict spe-
Amer. J. Bot. 80:
oO ^ч
T
doscorea (Dioscoreaceae).
9
Thorne, К. 19
ranges of a TIRA
— 1%
Major и іп E oe
Juart. Rev. Biol. 4
SEE Pp. | * М in pum A Ed-
of North America North of
xico, Vol. 1. O ыы Univ. Press, New York.
Tiffney, B. H. T Perspectives on the origin of the flo-
ristic similarity between eastern Asia and eastern North
America. J. Arnold Arbor. 66: 73-94,
Torrey, J. & A. Gray. 1840. Flora of North America, Vol.
Part 4. Wiley € Putnam, New Yor
Valiejo- — К. M., M. G. Pimenov, E. 1. Terentieva, S.
R. De , D. S. Pa ees & A. V. Troitsky. 1998.
Molecular. ae matics of e Using —
rDNA internal transcribed spa quences to reso
issues of evolutionary rel: dine Tol Zhurn. 83:
22.
torial € EU e,
к е
е К. G. 1972. Michigan E ar 1.
u onoc ols. — Inst. . Bull.
Vuilleumie г, В. 5. 1971. Pleistocene c — in же айпа
and flora Y South е Science 173: 771—780.
£
ا
=
pe aad rms
—188
Wake, D. B., С. Roth & M Wake. 1983. On the prob-
lem = — in — evolution. J. Theor. Biol.
101: —224.
Walck, : L., 5. N. Hidayati < N. Okagami. 2002. Seed
germination ecophysiology of the Asian species Osmor-
hiza aristata (Apiaceae): Comparison with its North
American congeners and implic pde dor oo of
types of dormancy. Amer. J. Bot.
998 eastern Aa * eastern
North American disjunct pattern: onis bom phylo-
genetic — Korean J. РІ. Taxon. 28: 63-81.
————. 1999. Evolution of eastern. Asian and eastern
North b. 'an disjunet parte rn in flowering plants.
Ann. Rev. Ecol. Syst. 30: 421—455.
~ Internal transcribed spacer phylogeny of
e astern North American disjunct Aralia sect.
—— (Araliaceae) and its biogeographic im-
plications. Int. J. Pl. Sci. 161: 959-966
u “2001. Evolution of eastern Asian—eastern North
American ЕТ disjunctions: A few additional
issues. Int t. J. РІ. Sci. 162: $117-S122.
& ы; 1996. Phylogeny and biogeog-
raphy of Panne | „ (the — genus, Araliaceae): In-
ferences from I Is sequences ¢ ue л "ar ribosomal DNA.
Molec. Phylogenet. Evol. € - 166-
. К. Jansen & К. — 1996. Evolution of
E — rn Asian and eastern North American disjunct
genus Symplocarpus (Araceae): Insights from chloro-
plast DNA restriction site data. Biochem. Syst. Ecol.
24: 735-747.
———, 5. Shi, K. Jansen & E. A. Zimmer, 1998.
Phyloge ny т ы of — sect. Aralia (Ar-
aliac eae). . Bot. 85: 866-87
Wood, € "1970. Some floristic relationships between
the onthe тп Appa 'hians and western North America.
Pp. 331404 in P. C. Holt (editor), The Distributional
History of the Biota i the Southern Appalachians, Part
Il: Flora. Research Division Monograph 2. Virginia
Polytechnic Institute and State University, Blacksburg,
Virginia.
. 1972. deal aan d ie phytogeography: The clas-
sical — :h to the study of disjunctions. Ann. Mis-
" 4.
ue Bot. Gard. E 107-12:
Wu, C.-I. € W.-H. 1985. Evidence for higher rates at
nucleotide ышы in rodents than in man. Proc
Natl. Acad. Sei. U.S.A. 82: 1741-1745.
Wu, 7. Y. 1983. On the significance of Pacific intercon-
tinental discontinuity. Ann. Missouri Bot. Gard. 70:
977-56 B
Q.-Y Crawford, A. D. Wolfe, Y.-C. Tang &
J.
. De E 1998. Mi d ч н а of
i har L. (Hippocastanaceae): ju cular phyloge-
netic pe rspec live. d 52: ou
Yokota aw . lida, | — 8. Tanifuji.
1989. m de otide s — nees of A: 5.85 rRNA gene and
internal transcribed spacer regions in carrot and broad
bean ribosomal DNA. J. Molec. Evol. 29; 294—301
A SYNOPSIS OF OCOTEA
(LAURACEAE) IN
CENTRAL AMERICA AND
SOUTHERN MEXICO!
Henk van der Werff
ABSTRACT
Ocotea is the largest genus of Lauraceae in Mesoamerica (Central America and Southern Mexico, i.e., the area
between the Isthmus of Tehuantepec in Mexico and the Panamanian—Colombian border) with 102 species — jd.
The most recent treatment of Mesoamerican Ocotea was published by Carolyn Allen in 1945. It included 3
it is curre ntly known
some brief notes on characters useful in their identification,
and is now se — outdated. A synopsis of the genus as
consisting of a lo species, synonymy, type Табынай,
and distribution rid country and altitu ide.
Key words: Central Americ a, Lauraceae,
Mesoamerica, Ocotea,
spec ies
from Mesoamerica is present d here,
southern Mexico.
Lauraceae are a large and ecologically important
family of trees and shrubs in wet tropical forests
from sea level to the tree line at about 3000 m. The
family is rare or lacking in areas with a pronounced
dry season. Lauraceae have the reputation of being
a difficult family, partly because of problems with
generic delimitation and partly because many spe-
cies are trees with inconspicuous flowers and there-
fore difficult to find and are infrequently collected
with flowers. A key to the genera of Lauraceae in
the Neotropics was published by van der Werff
(1991); this
characters but did not include fruit. characters.
key
makes it almost impossible to use because herbar-
key was based on floral and vegetative
Combining fruit and floral characters in :
ium specimens do not as a rule have both flowers
and fruits. By far the largest of the genera of Laur-
aceae in the New World is Ocotea Aubl., with an
estimated 300+ species in the Neotropics (van der
Werff, 1991) or 350 including the African and Mad-
agascan species (Rohwer, 1993). Ocotea is char-
acterized by its stamens with 4 locelli; these ar-
ranged in 2 pairs, unisexual or hermaphrodite
flowers, a lack of papillae on the stamens or tepals
in most species (papillae are present in a few spe-
cies, but these have the locelli clearly arranged in
2 pairs), and the tepals free, not basally united, and
falling off in old flowers or rarely persisting on the
cupule. In this way Ocotea is not very well defined.
and both van der Werff (1991) and Rohwer (1993)
suggested that Ocotea serves as the catchall genus
for species with 4-celled stamens that do not fit in
the other, better-defined genera of Lauraceae. The
Central American and Southern Mexican (from the
area between the Isthmus of Tehuantepec in Mexico
and the Panamanian-Colombian border, hereafter
referred to as Mesoamerica) species of Ocotea were
(1945), who included 33 spe-
cies in the genus. Since then, additional collections
ast revised by Allen
from Chiapas and Guatemala have led to the de-
scription of several new species by Lundell (1965.
1969, 1970, 1971, 1974a, 1974b, 1977, 1978). In
the more recent treatment of Lauraceae for the Flo-
ra Costaricensis (Burger & van der Werff, 1990) 42
species of Ocotea were recognized, including 7 new
species. Rohwer (1991) reviewed the marginal spe-
cies of the O. helicterifolia group, including the O.
heydeana group, and described four new species.
Van der Werff (1999) treated the core species of the
O. helicterifolia group and found four additional
new species. The study of recent collections of Me-
soamerican Lauraceae has resulted in publications
in which new Ocotea species have been or will be
described (Gomez-Laurito, 1993, 1997; Hammel,
1986: Lorea-Hernandez & van der Werff, 2002 in
Nelson, 1984; van der Werff, 1987, 1988a,
press:
1988b. 1996, 2001: Wendt, 1998; Wendt & van der
Werff. 1987).
SPECIES GROUPS IN OCOTEA
In a genus as diverse as Ocotea, one can expect
to find several distinct. species Rohwer
(1986) published an overview of the species of Nec-
groups.
! I thank n curators of B, BR, CAS, F, K, NY, US, and TEX for loans and Gordon McPherson for critically reading
the manuscr
? Missouri ‘Botanic al Garden,
ANN. MISSOURI Bor.
P.O. Box 299, St. Louis, Missouri 63166-0299, U.S.A. henk.vanderwerff@mobot.org.
GARD. 89: 429—451. 2002.
Annals of the
Missouri Botanical Garden
Table 1. Ocotea species with unisexual flowers in the
Mesoamerican region.
Table 2. Species of Ocotea subg. Dendrodaphne in the
Mesoamerican region,
Ocotea adela
Ocotea puberula
Ocotea rubrinervis
ex Rottb., Ocotea,
Nees, and Rhodostemonodaphne Rohwer € Kubitz-
tandra Roland. Pleurothyrium
ki, largely based on a study of type specimens. He
recognized a number of species groups in Ocotea
based on characters of flowers and fruits. Among
the Mesoamerican species the following groups can
be recognized, starting with the best-defined
groups:
l. Species with unisexual flowers. This group
comprises almost half the species of Ocotea and is
best represented in South America. In Mesoamerica
one finds 12 species of this group (Table 1). Most
of those are also present in South America, and
only two are endemic to Mesoamerica (O. adela van
der Werff and O. atlantica van der Werff). Rohwer
(1986) divided this group into numerous smaller
groups, based largely on the South American spe-
cies. The Mesoamerican species can all be confi-
dently assigned to these smaller groups, but listing
these would be too detailed for this synopsis.
. Ocotea subg. Dendrodaphne (Beurl.) Mez.
This group is characterized by a combination of the
following characters: inflorescences in the axils of
bracts near the tips of the twigs; cupules with a
double-rimmed margin; flowers with spreading te-
pals; and nearly sessile, tongue-shaped stamens
with a sterile tip and a papillose indument. This is
a small, but well-defined group consisting of about
10 species, with 4 species in Mesoamerica (Table
2
The Ocotea helicterifolia group. This group
shares some characters with the species of Ocotea
subg. Dendrodaphne and can be recognized by the
following combination of characters: inflorescences
in the axils of normal leaves; cupules with a single
margin; flowers with spreading tepals and stamens
subsessile or infrequently with a distinct filament,
sometimes tongue-shaped, densely papillose and
with a sterile tip or more commonly rectangular,
without a sterile tip and very sparsely papillose.
Ocotea dendrodaphne
Ocotea klepperae
Ocotea morae
Ocotea veraguensis
This group has been recently revised by Rohwer
(1991), who studied O. heydeana (Mez & J. D.
Smith) Bernardi and its allies, which have glabrous
or very sparsely pubescent leaves, and O. sinuata
(Mez) Rohwer and its allies, which have densely
pubesc ent leaves and densely papillose stamens
(1999), who studied O. helicterifolia (Meissner)
Hemsley and its allies, characterized by pubescent
leaves and stamens with few papillae, without a
sterile tip. This group is strongly centered in Me-
soamerica, with only one species in northern South
America, and includes about 32 species (Table 3).
The Ocotea insularis group. This group is
characterized by two seemingly insignificant char-
acters. One is the pubescence of the inner 3 sta-
Table
3. Species of the Ocotea helicterifolia group in
the Mesoamerican region.
l. Ocotea acuminatissima
2. Ocotea arcuata
3. Ocotea —
4. Ocotea betazer
E Ocotea botrar
. Ocotea bonrgeauaiana
. Ocotea brenesi
" ( s congregata
2.
=:
9. Ocotea corrugata
10. Ocotea gordonii
11. Ocotea helicterifolia
12. Ocotea heydeana
13. Ocotea —
I4. Ocoted lenti
15. Ocotea agni
16. Ocotea тойс
17. Ocotea n
18. Ocotea pa
шаса
] 9. Ocotea — * hrosorum
25. Ocoted rubriflora
26. Ocotea sinuata
27. Ocotea tonii
28. Ocotea valeriana
29. valerioides
30. Ocotea verticillata
31. Ocote
—
^ad
>
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
Table 4.
Mesoamerican region.
Species of the Ocotea insularis group in the
1. Ocotea atirrensis
2. Ocotea austinii
3. Ocotea barbatula
Ocotea chi |
5. Ocotea contrerastt
6. Ocotea endresiana
ea whitei
16. Ocotea viridiflora
mens; these stamens have a patch of white hairs on
the side facing the pistil at the junction of the fil-
ament and the anther. The second character 15
found in the inflorescence: the secondary, tertiary.
and further branches of the inflorescence are flat-
tened, with the axis of flattening of the tertiary
branches forming a 90° angle with those of the sec-
ondary branches, those of the quaternary branches
forming a 90° angle with the tertiary branches, and
so on. In other characters this group conforms to
the general Ocotea pattern: inflorescences in the
axils of leaves; cupules with a single margin: flow-
ers with erect to half-erect tepals and stamens with
a distinct filament, not papillose, and without a
sterile tip. This group is best represented in Me-
soamerica, with one or a few species in northern
South America. I would place 16 Mesoamerican
species in this group (Table 4), including O. me-
ziana С. К. Allen and its allies, which differ from
the O. insularis group mainly in the light green
drying leaves with raised reticulation and often yel-
lowish major veins. My concept of this group
considerably wider than that of Rohwer (1986).
larger
un
eyond these reasonably well-defined,
groups one can establish a larger number of poorly
defined, small groups, but doing so is beyond the
scope of this synopsis. Too many species would be
transitional between such groups, and they would
contribute little to our understanding.
If Ocotea is the genus to which all species that
cannot be placed in other, better-defined genera of
Lauraceae are assigned, one may ask if Ocotea, as
currently accepted, is monophyletic. The answer to
this question can be drawn from a recent DNA-
based phylogeny of Lauraceae (Chanderbali et al.,
2001). In this study 22 species of Ocotea, repre-
senting several different groups, were included.
Those results show clearly that Ocotea is not
monophyletie genus, but that the following clades
can be recognized:
1. The 9 species with unisexual flowers formed a
clade together with Endlicheria Nees and Rho-
dostemonodaphne, both genera with unisexual
flowers.
2. The 3 species of the O. helicterifolia group
formed a clade.
The 2 species of subgenus Dendrodaphne
formed a clade.
4. The 4 Old World species formed a clade.
5. Three species with perfect flowers (including O.
insularis (Meissner) Mez) formed a clade with 3
Aniba species and 1 Aiouea species.
6. A recently described species from Cameroon, O.
ikonyopke van der Werff, was part of a clade with
Cinnamomum species and was probably incor-
rectly placed in Ocotea.
These clades are well separated from one anoth-
er, and maintaining Ocotea as a monophyletic ge-
nus would require a transfer of several quite dis-
tinct genera (such as Nectandra, Pleurothyrium,
Aniba Aubl., Licaria Aubl.) into Ocotea, a solution
that Edo not find acceptable. Hence Ocotea should
be considered a polyphyletic genus. Mez (1889)
recognized 198 species in Ocotea and divided the
genus into 4 subgenera: subg. Hemiocotea Mez,
with 1 species, now placed in the genus Urbano-
dendron Mez; subg. Dendrodaphne, 8 species, with
the same circumscription as it currently has; subg.
Mespilodaphne (Nees) Mez, consisting of all re-
maining species (51) with perfect flowers: and subg.
Oreodaphne (Nees) Mez, including all species (138)
The DNA study supports
the monophyly of subgenera Dendrodaphne and Or-
eodaphne, but not of subgenus Mespilodaphne. Mor-
with unisexual flowers.
phological characters for a division of subgenus
Mespilodaphne in monophyletic groups have not yet
been found
This synopsis is based on a treatment of Ocotea
for the Flora Mesoamericana and includes the 102
species found from the Isthmus of Tehuantepec in
Mexico south to the border between Panama and
Colombia. Mexican species only known from the
area north of the Isthmus of Tehuantepec are not
included. Altitudinal distribution for species also
occurring outside of Mesoamerica is based on Me-
soamerican specimens and does not reflect their al-
titudinal distribution outside this area.
Annals of the
Missouri Botanical Garden
KEY TO THE SPECIES OF OCOTEA IN MESOAMERICA
1: 3(12).
13
14(13).
ee
2221).
22".
23(20).
23'
Stamens tongue-shaped, with a distinct, sterile tip, moderately to densely papillose; filaments lacking
E VEE lii ec P
а. ns not tongue-shaped, usually т rectangular, without a ste rile tip, not inni or rarely with papillae
on the margin; filaments usually Present ы A — l
Lowermost pair of lateral veins ihi conspicuous Lcd domatia — 41.0. holdridgeiana
Leaves without conspicuous poc * ‘t domatia
‘Twigs and usually the leaves densely pubescer ent
‘Twigs and leaves glabrous or ance m — {
pee clustered— a - 99. O. verticillata
Leaves evenly distributed along the twig - 5
Lower leaf surface completely covere d pe the indument, the hairs short and cure е —
66. О. pharomac hrosorum
Lower leaf surface not completely covered by the indument, but partly or — completely visible )
Leaves elliptic, to 6 cm wide; petioles 5—9 mm long Tm 46. O. klepperae
Leaves obovate or obovate-elliptic, 5—17 cm wide; petioles 1. 54 cm long 7
Inflorescences paniculate-cymose; pistils с pubescent п 84. nuata
Inforescences racemose; pistils glabrous m 13. 0. urina
‘Twigs hollow | —
Twigs solid ү
Shrubs or small trees to 8 m; leaf bases plane, not revolute; inflorescences to 15 ст long; fruits ca. 2
X l em; сирше ca. | em diam., cup- to bowl-shaped, very weakly канон e е
1. 0.5 mm wide; widespread |... 1. 0. deboli
Troes to 20 m; leaf bases slightly revolute; infloresc 'ences to 5 em long; fruits ca 5 E m cm, cupule
ca. 5 em diam., bowl-shaped, with a conspicuous double margin, this ca. 1 em de Costa Rica
E TAANI E A инал O A 0. morae
— ‘ences about as long as the leaves, пената pubescent; Teal bakea — not гак pus 2
em, cupule to 1.2 em diam., with a double margin, the outer margin ca. 2 mn wide | widespr reac
a
— . 0.1 —
Inflorescence ‘es much shorter than the le saves, de nsely pubescent; leaf bases slightly — fruits
5 X 3.5 em, cupule ca. 5 em diam., with a conspicuous double margin, this ca. 1 em wide: — Rica
‚ O. morae
Lower leaf шк е complete sly hidden by the indument — а. 1:
Lower leaf surface glabrous or partly covered by the indument 5 17
Flowers ш айди leaf base inrolled and decurrent along the entire length of the petiole; —
lines visible on the lower leaf surface — 16. О. calophylla
Flowers perfect; leaf base plane or if inrolled and decurrent, most of the petiole distinct; vernation lines
not present — — —
Leaf M inrolled and shortly decurrent on the petiole | 83. O. salvinii
se plane and not decurrent 1
indias nt of the lower leaf surface pale brown or whitish у. 15
Indument of the lower leaf surface reddish brown zm 16
Domatia absent; flowers 6 mm diam., tepals 2-2.5 mm long — 0000 82. O. salvadorensis
Domatia present as axillary tufts of hairs; flowers 4 mm — tepals 1.5 Ер long . 43. O. iridescens
Receptacle glabrous inside; tepals papillose on the inner surface; leaves 9-16 cm long 81. O. rufescens
Receptacle appressed pubescent inside; tepals pubescent on the inner Ре 5 leaves 7-10 ст long
see discussion iul 82. O. salvadorensis
Twigs hollow, often inhabited by ants me 18
Twigs solid 0 00000000 n 19
Inflorescences and flowers de ansely brown-tomentellous; leaves s broadly obovate, coriaceous, with raised
reticulation 44. O. jefensis
Inflorescences and flowers glabrous or sparsely puberulous; leaves (narrowly) elliptic or obov: ate, ы
chartaceous, reticulation immersed or raised —.— ). atirrensis
Flowers unisexua : — ИРЕЙ 20
Flowers perfect o — 000000 - 30
Lower leaf surface with erect hairs, disce sable Bin.) Ka ce I RR RR 21
ower leaf surface glabrous or with appressed hairs, hairs not discernable to the touch — 23
Indument completely covering the surface of the twigs 0000000000000 22
Surface of the twigs partially visible between the hairs 51. О. mac —
e slightly decurrent on the petiole; laminae and pe ticles. not sharply offset; tertiary Еи or
e lower leaf surface scalariform and clearly raised; indument on young twigs s agey O. fendleri
В. ase of laminae obtuse, clearly differentiated from the petiole; tertiary venation on lower * —
carcely raised and not scalariform; indument on young twigs very short, « pe lica
Twigs sharply angular; upper leaf surface with raised, minute reticulation 7 a — **
‘Twigs rounded or slightly angular; upper r leaf surface without raised reticulation . or vilis coarse retic
lation
—
-
Volume 89, Number 3 van der Werff 433
2002
Synopsis of Ocotea
24(23).
F
25(24).
ES
28(27).
28".
29(28).
29'.
30(19).
30'
31(30).
31
32(31).
36(: 35).
36
37 (ЗО).
A.
38(37).
39(38).
39'.
40(35).
43(42).
43'.
Hi 10).
A x 14).
45
э 15).
25.
26(25).
Pistils pubescent; cupules plate- like, with a double margin |... = . О. —
Pistils glabrous; cupules various, but not plate-like and with a double r margin .-
Pit domatia present in the axils of the lateral veins of some leaves, the domatia not pubese e d
PO TD FRE . О. oblonga
ри domua absent, but axillary tufts of hairs sometimes present 26
Axillary tufts of hairs present on the lower leaf surface; tepals persistent on the cupule; oil glan ds
readily visible on the upper leaf surface as black dots 0 80. O. rubrinervis
Axillary tufts of hairs lacking: te pals deciduous or persistent in fruit; oil glands not or scarcely visible
Twigs densely and minutely gray appressed pubescent, the hairs small and scare ely ااا visible;
indument on the inflorescences and pedicels notably denser than on the flowers ). —
Twigs sparsely pubescent, individual hairs visible, or glabrous: inflorescences and — UE 0
density of indument on flowers and inflorescences about the same
PARA AREA ee ER 7. О. cernua
29
Piia of Ihe. outer 6 stamens free; inflorescences ети idus ‘rs sparse els ae ent ¢
Leaves 7-25 cm long, the tertiary venation slightly raised on the upper surface; lateral veins —
on the upper ине же Р ОЕТИРНОРИ ЕН 7: ). рибе rula
Leaves 4-11 cm long, the tertiary venation imme sed on the upper surface; lateral veins гань im-
pressed on * upper surface |... ). ). adela
Lower leaf surfaces and young twigs 5 with pre 'edominantly erect hairs, these discernable to the touc ^s 31
Lower leaf surfaces and young twigs glabrous or with predominantly appressed hairs, rarely with sparse
erect hairs, but these not discernable to the peek
Leaf bases decurrent on the petiole and usually inrolled 0
Leaf bases not decurrent on the petiole, usually flat 35
Young twigs appressed pubescent, the indument moderate or sparse and part of the surface alway
visible; inner surface of tepals glabrous or nearly so 00 25. О. —
Young twigs with erect or ascending hairs, the surface (nearly) completely covered; inner — 'e of
tepals pubescent ишсен —— З.
Leaf bases al recurved: lateral veins 9 to 12 pairs 0 86. O. stenoneura
Leaf bases scarcely if at all recurved or inrolled: lateral veins 4 to 8 pair 7 34.
Indument on lower leaf surface predominantly erect; leaves 10-20 cm E А 37. O. hartshorniana
Indument on lower leaf surface predominantly appressed; le saves 5-12 em long —_ 57. O. montei erde nsis
Outer surface of tepals densely pubescent, the surface completely — by the indument if inne
tepals with less indument on the upper half, then at least outer 3 tepals densely pubescent 30
Outer surface of all tepals glabrous or variously pubescent; if variously pubescent, de surface of the
Flowers 10—14 mm diam и 33. О. gomezil
Flowers less than 8 mm diam 37
Inflorescences few-flowered, racemose or with some lateral Grane hes ending in a cyme; le aves сопасе‹
cC AI AAA . pse opal
Inflorescences many-flowered, the lateral branches several times divided: le saves coriaceous or chart:
. — — 38
— Е 7 — 3. O. amplifolia
Petioles up to 15 mm Іор 39
Petioles ca. 30 mm long
g —
Leaves coriaceous, densely ferruginous pubescent on the lower surfac е; ‘cupule with entire margin: leaf
apex obtuse or very shortly Een Tot LIS A RE A oe Pine EE 23. О. darcyt
Leaves chartaceous, moderately to sparsely golden brown abuso ent on the lower surface; tepals |
56. О. mollijalia
sisting on the cupule; leaf apex acuminate
Inflorescences racemose or rarely with a few lateral cymes os
Inflorescences panic "ulate-cymose 05 — 44
Midvein, lateral veins, and tertiary venation clearly impressed (leaves rugose- bullate) — 22. О. corrugata
Venation immersed or raised, not impressed (leaves not rugose-bullate) —— — 42
Outer surface of tepals glabrous: anthers sessile 0000000000 sl 34. 0. gordonii
Outer surface of tepals (sparsely) pubescent: stamens with filaments Y or more of the bli of the
STI AA ыы RE EDT RI PRACT VU —
Hairs on the lower surface of leaves asce nding. covering most of the lamina; indument PLAY eee
авуз eee 55. O. mollicella
isis on iile bus: en] suis e erect, most of the lamina eines indui brown or ү дитин
ge арн A ھی ت ت ی A o ТАК O DUPDUTGO
Leaves clustere d. me OM cat ETE O 45
Leaves alternate, evenly distributed along Ше twigs — ee AN ee a n — 7
Outer surface of the tepals glabrot — — = СТО ТЕНЧ 14. 0. bourgeauviana
Outer surface of the tepals генен TER NIE NER AEN _.. 46
Petioles to 6 mm long: leaves 15-25 X 5-6 c m; indument el twigs ye -llow-brown __ А 91. O. tonii
Petioles at least 10 mm long; leaves 10-15 X 6-7 em; indument of twigs brown 20. О. congregata
Annals
— — Garden
47(44).
7!
a.
4847).
" 18).
50 49).
50
5
5 51 (48).
51”.
5530).
55'.
56(55)
ә
60(59).
60”.
61(59).
ol’.
62(61).
62’.
63(62).
03".
64(63).
Flowers at anthesis 2-3 mm diam.; tepals erect to half-erect; inner surface of tepals and stame
glabrous f ›2. ( ШАТ
Flowers at anthesis 5-10 mm diam.; tepals spreading or half-erect; inner surfac e of tepals and/
stamens partially papillose or vani nt
side
Receptacle pubescent in 49
Receptacle glabrous inside — < - Ё 51
Leaves elliptic, to 15 em long : e 12. O. betazensis
Leaves obovate, 20-40 cm long . a = 2 90
Indument completely covering young twigs and inflorescences 95. O. valerioides
Surface of twigs and inflorescences visible between the indument _ — Е . lentü
Outer surface of the tepals pubescent E : — | " 64, 0. =
Outer surface of the tepals glabrous — сос. =
Inner surface of the tepals pubescent; filaments distinct, c ca. Y the length of the anthers — E |
Inner surface of the tepals glabrous; filaments not evident = Е m 54
Leaves to LO cm long, the tips obtuse, acute or shortly acuminate O. praetermissa
Leaves 10-20 cm long. acuminate or gradually narrowed into a slender tip . 4 bourgeauviana
Surface of young twigs completely covered by the indument; cupule cup- shaped 94. O. valeriana
Surface of young twigs partially visible between the indument; cupule shallowly howl-shaped or plate-
like 38. helicterifolia
Leaf bases inrolled and/or decurrent on the petiole or reflexe d, or leaves subsessile with the — weakly
recurved; free petiole usually absent or very short — m - 56
Leaf bases not inrolled or decurrent on the petiole; petioles usually distinct 65
Flowers densely reddish pubescent; terminal buds glabrous or fine ly appresse d pubesc ent distally
— 90. O. tonduzti
Flowers gray pubescent or Кини te mina buds uniformly pubescent or "nd glabrous — 57
Leaves usually 30-50 em long, the apex usually rounded; domatia lacki E rivularis
qi usually less than 25 em long, apices obtuse or rounded; nee ie ‘king or —
eaf bases reflexed, the lobes frequently covering the midrib; flowers glabrous ... — 27. О. н жа ын
Leaf зеби decurrent or inrolled, but not reflexed; flowers usually puberulous 59
Flowers 4-5 mm diam E - . - — 60
Flowers 2.5-3 mm diam 61
Outer surface of the tepals densely whitish pubescent, the surface complete ly cove — vernation lines
visible on the lower leaf surface; receptacle (deeply) — Costa Rica, Pan v
: al aucosericea
Outer surface of the te de moderately pubescent, the — е pab visible; vernation — not visible
8. (
rec — ‘le bowl-shaped; Mexico, Guatemala О. chiapensis
Leaves 5-10 em wide, obovate, glabrous or with a few appressed hi hairs on the lower leaf surfac e; tertia
venation not or scarcely raised on upper leaf surface ). insularis
Leaves to 5 em wide, elliptic or oblong, appressed pubesc ent or ‘glabrous | on the lower leaf surface
tertiary venation imme sed or raised
Leaf base cuneate or acute, moderate Чу or weakly reflexed, not decurrent on the petiole; leaves sub-
sessile — 42. O. insularis s.l.
Leaf base decurrent on the petiole, usually inrolled; leaves petiolat ite 63
Tertiary venation raised on the upper surface: leaf tips obtuse; leaf bases inrolled and decurrent
vie austinit
Tertiary venation not or rare ly raise d on the upper — if raised, leaf tips acute or acuminate and/
or leaf bases decurrent, but not inrollec
Apparent petiole due to decurrent leaf bases, 4—5 cm long; flowers sparsely — 70. 0. produci
Apparent petiole to 3 em long, usually shorter; indument of flowers various 10 whitet
Tepals papillose or partially papillose on the inner surface; te pals glabrous. or sparsely — nt
outside; tepals spreading at anthesis; tepals usually longer than З mn 0
Tepals glabrous or pubescent on the inner surface; outer surface — or variously pubescent: tepals
usually erect or half-erect at anthesis; — usually less than 3 mm long А 9
Hairs on the lower leaf surface erect (check along midrib and lateral ve ins) Е Е 67
Hairs on lower leaf surface appressed or lacking
Inflorescences racemose; young twigs densely pubescent with minute, erect hairs 67. О. pittieri
Inflorescences panic ulate-c 'ymose; young twigs moderately to densely appressed pubese ent
›. О. rhytidoirie ч
Stamens with lum nis at least as long as the ене outer stamens с used inward 5. O. pausiaca
Stamens with filaments less than half as long as the anthers; outer stamens straight Loss. 6
Pit domatia present in the axils of the lowermost late ice veins — 4. O. arcuata
Pit domatia lacking, axillary tufts of hairs sometimes present — 70
Inflorescences densely and minute dy brownish pubescent, the surface е cove етей « or nearly so .. с. di
ose ‘ences sparsely or moderate ‘ly pubescent, the surface largely visib » 72
Flowers 6-7 mm diam.; widespread, from southern Mexico to Panama | 79. O. rubriflora
Flowers 10-12 mm diam.; Osa Peninsula in Costa Rica EA е | 50. О. macrantha
Leaf margin at the base (narrowly) reflexed or inrolled - . А 3 — 68. O. platyphylla
Volume 89, Number 3 van der Werff 435
2
002
Synopsis of Ocotea
-l
2 N
— гь
а
І
لح لد لد لد لد لد لد له 1
b de ЕЧ 3
79.
80(79).
80”.
аш
in
82
А 2" ,
—
—
А p
84”.
8584).
85'
86(83).
80".
91(86).
87'.
88(87).
88'
8919).
89”.
4 89).
(90).
92(89).
92’.
93(92).
93'.
Leaf margin at the base flat or nearly so
ES
w
Inflorescences paniculate, with at least a few e flowered lateral cymes, rarely racemose but then with
flowers 6.5 mm or more in diameter and leaves drying greenish 00 0 0 — 1
Inflorescences racemose; leaves drying dark green or flowers * io 5 mm X — 77
lama nt on twigs brown and + istent; indument on peduncle moderately dense 15
Indument on twigs pale and becoming quickly sparse with age; indument on pe * le rather sparse
ateral veins 6 to 9; Guatemala and С iuis
—
í
— 52. 0. —
Lateral veins 4 to 5; Costa AA 102. Ocotea sp. A
Flowers 4.5—5.5 mm diam.; lowlands — — 9. О. жоло
Flowers 6.5-9 mm diam.; montane forests, — above 900 m o 40. O. heydeana
Indume ‘nt on young twigs dense and consisting of short, erect hairs 00 _ 67. O. pittieri
— on young twigs of variable density and consisting of appressed. peu somewhat ascending
ha
Rec ae le pee pubes ent inside: Costa Rica
Rc PS 15. O. brenesú
Receptacle sparsely or very sparsely pubescent inside; Mexico, Cuán ala _ 1. O. acuminatissima
Lower leaf surface with pit domatia, cavities with a small slit-like or rounded же е, in the axils of the
basal lateral veins or along the lower lateral veins; orifice — glabrous, rarely with a fringe of hairs
A A sss 80
ere leaf surface without pit denda: sulla tufts of PONE in a , small depression sometimes present
A E A E E E aes 89
Domatia at least s once, |, often — times, their diameter away е from. the midrib; three or more р
domatia present in each leaf; domatia with a fringe of hairs cc 10. Vh
Domatia adjac ent to or less than their diameter away from the midrib; domatia present in the pea
several pairs of lateral veins or only in the axil of the basal pair; domatia usually glabrous 8l
Flowers sparsely appressed pubescent: inflorescences moderately appressed pubescent |. 36. О. педаг
Flowers glabrous; inflorescences glabrous or nearly so s 2
Leaves tripliveined; domatia only present in the b 'rmost pair of lateral veins 96. О. —
Leaves pinnately veined; domatia often present in the axils of more than one pair rof lateral veins or
along the lateral VES a AAA ا v UE нд» 83
Terminal buds glabrous — ——— 84
Terminal buds pubescent ici. retratos | - 86
Leaves broadly elliptic, 17-27 X 8-15 BS GI. ee НИЯНЫН — ‚ 29. O. euvenosa
Leaves elliptic, A MM 85
Tepals е at anthesis; leaves. n n des, vellowish green, the validi often ома in aes than
the laminae ne 100. O. viridiflora
Tepals erect or half. erect al anthesis: [en s, when dry, dark $ green to o Bhi dus the midrib cone ue Is
with the laminae o — 89. O. tenera
Leaf apices obtuse to нн leaves coriaceous: doehas conspicuous, restricted to the — of the
— O. pullifolia
apices acute; — ee teous to coriaceous: T, coriaceous, domatia A imd found e the
lowermost pairs of lateral vei
af à
lowermost lateral veins |...
Tertiary venation on мун le af "en e immersed or nearly so; hi ‘aves coriaceous to — каа eous,
dark green wher
=
we
о O RA A e : 5. Jorge-esc barii
Tertiary venation on vd leer — e raise d: Je aves chartaceous, bled k or velldiial green when dry
DONDE CON RR IM = u 88
Inflorescences to 20 cm long; mm "s hac k when dry, the ите! very vais E. 11. O. bernoulliana
Inflorescences to 10 em long: leaves yellowish green when dry, the main veins often lighter in color
than thé amina sees unt caen adea EE . 54. О. meziana
Terminal buds deci тере times a few hairs present along the margins of the bracts das the
s)
TS ER E 90
Terminal buds pubescent лыдан с чайнады села л E рене асс оди 92
Flowers and inflorescences alata »us — 98. О. verapazensis
Flowers and distal parts of inflorescences sparsely or moderately, minutely ole scent 91
Leaves 15-35 X 7-15 em, leaf base rounded, rarely obtuse or subcordate: tepals erect or ne ee E
anthesis 22000000 85. 0. —
Leaves 10-15 X 4—5 cm, leaf ‘hase acute; tepals spreading at anthesi 53. O. matudat
Twigs densely pom ent, the surface — tely covered by the erect or ascending hairs
el و cle ee cs ا ا چ 21. 0). contrerasti
Twigs g — pen ми ent or, if densely pubescent, the hairs appressed 93
Leaves, when dry, yellow-green, rarely darker green: tertiary venation forming a raised retic iln a on vba
lower leaf surface: midrib and lateral veins usually lighter in color than the —— leaf tis 94
Leaves, when dry, dark green: tertiary venation not forming a raised reticulum on the lower leaf seg
major veins concolorous with or darker than the leaf tissue B ER . 95
436
Annals
la — Garden
94(93).
94
95(93).
95'.
id 5).
—
97'.
98(97).
98".
99(98).
99”.
100(96).
100".
101(100).
101”.
102(101).
102'
103(102).
103’.
104(103).
105100)
— 104).
106’.
l. Ocotea acuminatissima (Lundell) Rohwer,
Bot. Jahrb. Syst. 112: 379. 1991. Phoebe ac-
uminatissima Lundell, Contr. Univ. Michigan Distribution.
Herb. 6: 19. 1941. Cinnamomum acutatum from 900 to 2700 m.
Twigs and leaves glabrous or with a few appressed hairs 47. O. laetevirens
Twigs and major veins on the lower leaf surface ж er pubescent with very short, erect or asc alk y
hairs; indument on the lower leaf surface similar, but sparser |. ). rovirosae
Tepals to 1 mm long and inflorescences moderately to sparsely pubesc ent, the hairs asc scending e
-]
zl
л
-
). mu dom
Tepals at least 1.5 mm long; if rarely shorter (1.3-1.5 mm), then inflorescences very sparse ly pubescent
with appressed hairs
Inflorescences densely pubescent, the: surfac e large ly: or completely covere ad by the indúument - 97
Inflorescences sparsely or moderately pubescent, the surface largely or entirely visi 100
Leaf apices rounded or obtuse; axillary tufts of hairs lac кш on the lower leaf surface
=
Leaf apices acute or acuminate; axillary tufts of айа present on the lower ЗГА ас
eaves obovate, to 4 em wide, drying green; twigs, when — alate or — ne Jover leaf
surface sparsely pubescent, the hairs short, erect and inconspicuous .€
Leaves elliptic or oblanceolate; when widest above the n
lower leaf surface sparsely appressed pubescent or glabre
eaves to 2.5 em wide, oblanceolate or narrowly obovate- valine. drying blac К; — ca. 15, mm long
9. O. chry —
98
eucuneata
— ane blac k; twigs — or angular;
О. кш
— 2.5-8 cm wide, elliptic c or narrowly у elliptic --obovate, drying green to dark green; ра! a. 2
nm lon 93. О. —
Young SO densely pubescent, the hairs ere ct; infloresce nces 15-20 cm long; domatia lacking
a ). subalata
Young twigs appressed pubescent or — ‘glabrous: if with some erect Tes беле — short
and their orientation scarcely visible; inflorescences usually shorter than 15
em; domatia, as axilla ary
tufts of hairs or shallow pits, often present
101
Leaves lanceolate; domatia consisting of shallow pits, these with a pubesc ent or sometimes glabrous
margin ). effi pes
Leaves elliptic, ovate or elliptic-obovate: domatia, if | present, consisting of axillary tufts of hairs 02
Leaves 13-25 X 4—8 cm, the tertiary venation raised on the lower surface 35. КОТ
— generally smaller than 12 X 5 em; if occ asionally to 15 cm long, the tertiary ve nation immerse
n the lower surface 103
— elliptic to oblong, the apices rounded or blunt; domatia consisting of tufts of hairs, : — the
own diameter away from the midrib herbert
Leaves ovate, obovate or e lliptic, the apices acute or acuminate; domatia, when present, in [iod Wr.
the lateral veins close to the midrib - —
Leaves ovate, distally tapering into a long and slender apex — re
eaves broadly or narrowly elliptic, without a long, slender apex =
Domatia present as axillary tufts of hairs; outer surface of tepals glabrous
Domatia absent; outer — е — tepals sparsely pubescent
Flowers glabrous or nearly
without magnification
63. 0. nu
‚ O. strigosa
; leaves › drying dark green; domatia (as axillary tufts sol hairs visible
75. O. truncata
Flowers pubescent, the indument covering the tepals almost completely; leaves drying green to
green; domatia, when present, not visible without magnification
^^
8
dar
‚ O. racemiflora
pressed indument on the lower leaf surface (or
leaves almost glabrous).
Mexico (Chiapas), Guatemala,
Kostermans, Reinwardtia 6: 20. 1961. nom.
nov. for Phoebe acuminatissima Lundell. .
— Өр IRA: 2. Ocotea adela van der Werff. Novon 11: 501.
PYPE: Mexico. Chiapas: Matuda 2107 (holo- mo 44 к 5 a
. . )О1. TYPE: Panama. Prov. Panamá: Cerro
type, MICH not seen; isotype, NY!).
Jefe, Croat 13049 (holotype, MO!).
Phoebe чиси лр, Lundell, Contr. Univ. Michigan
: : 14 2. TY
er
31
Phoebe — Lundell, Wrightia 1: 151. 194. E
namomun — — — ll) a Rein-
wart
A variable species characterized by its racemose
).
YPE: Mexico. Chiapas: Matuda An inconspicuous species best recognized by its
sed p not seen; isotype, | : unisexual flowers, leaves with slightly impressed
—
zu
lateral veins, and shallow cupules with persistent
tepals. It can be confused with the widespread Oco-
dtia 6: 23. 1961. TYPE: Mexico. Chiapas: Ma- hos (News): Mes. Ми iut Sheela haran
tuda 5140 tee, i not seen). бе сетш VANEGE ез, DML tnar эрестев паз orten
nodding flowers, immersed (not impressed) lateral
veins, and does not have persistent tepals on the
inflorescences, papillose inner surface of the tepals, — cupule.
spreading tepals at anthesis, and the sparse ap- Distribution.
Panama, from 500 to 900 m.
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
3. Ocotea amplifolia (Mez & Donnell Smith) van
der Werff, Novon 11: 510. 2001. Phoebe am-
plifolia Mez € Donnell Smith, Bot. Gaz.
(Crawfordsville) 19: 261. 1894. Cinnamomum
amplifolium (Mez & Donnell Smith) Koster-
Reinwardtia 6: 20. 1961. TYPE:
temala. Depto. Quiché: Heyde & Lux
(syntype, GH!).
Gua-
3035
mars,
Only known from the type collection. The long
petioles (3 em long) and the dense indument on the
twigs, inflorescences, and flowers are distinctive.
Distribution. Guatemala, 3000 m.
4. Ocotea arcuata Rohwer, Bot. Jahrb. Syst. 112:
380. 1991. TYPE: Panama. El
Road, Mori et al. 6883 (holotype,
Llano-Carti
MO!
A rarely collected species, best recognized by its
flowers with spreading tepals with papillose inner
surfaces, almost glabrous leaves with pit domatia
in the axils of the lowermost lateral veins, and the
loop-connected lateral veins.
Distribution. | Panama, from 300 to 550 m.
э. Ocotea atirrensis Mez & Donnell Smith, Bot.
Jahrb. Syst. 30, Beibl. 67: 18. 1901. TYPE:
Costa Rica. Donnell-Smith 4930 (syntype.
US”.
Ocotea — C. К. Allen, J. Arnold Arbor. 26: 345
19 TYPE: Pam. Coe lé: Allen 1211 (holotype,
F not seen; isotype
Ocotea ке, Mes [o — Syst. 30, Beibl. 67:
19. 190 : Costa Rica. Talamanca, Pittier &
Tonduz, in * inst. ge езйн Costarica 9172
9179 (syntypes, !).
Ocotea pentagona Mes, o — Syst. 30, Beibl. 67:
17. 1901. TYP
a Rica. Rio — Tonduz
in herb inst. * -geogr. Passus 7613 (syntype.
BR!); prope San Mateo, Biolley 7106 шо BR!):
without locality, Tonduz 8 = Pittier & Durand
3362 (syntype, BR!).
Ocotea wedeliana €. K. Allen, J. Arnold Arbor. 26: 339.
1945.
TYPE: Panama. Boc 'as del Toro: Cooper 539
(holotype, F not seen; isotype, GH!).
A broad concept of this species is accepted here:
distinctive are the hollow twigs, stamens without a
sterile tip. and small cupules. For a different view-
point, see van der Werff (1988a). Only a few Ocotea
species in Mesoamerica have hollow twigs: Ocotea
Jefensis van der Werff differs in its tomentellous
flowers and leaves with decurrent,
while O. dendrodaphne Mez has tongue-shaped sta-
inrolled bases,
mens with a sterile tip and a double-margined cu-
pule.
Distribution.
ama, from 50 to 1200 m.
Nicaragua, Costa Rica, and Pan-
Novon 11:
Za-
6. Ocotea atlantica van der Werff.
504. 2001. TYPE: Honduras. Atlántida:
mora 1744 (holotype, MO!).
Readily recognized by its small, unisexual flow-
ers and the erect pubescence on the lower leaf sur-
face and twigs; the indument obscures the surface
of young twigs.
Distribution.
Rica in forest on the Atlantic slopes from 50 to
100 m.
Honduras, Nicaragua, and Costa
7. Ocotea aurantiodora (Ruiz € Pavón) Mez.
295. 1889.
Laurus aurantiodora Ruiz € Pavón, Fl. Peruv.
1. 1. 349. 1804. TYPE: Ruiz & Pavon
n. (B not seen).
—
Jahrb. Kónigl. Bot. Gart. Berlin 5:
Peru.
Distinctive are the clearly angled twigs, the
small, raised reticulation formed by the tertiary ve-
nation on the upper leaf surface, and the unisexual
Ocotea longifolia HBK,
used for South American specimens, is probably a
flowers. a name widely
synonym.
Distribution.
much of South America, from 0 to 600 m.
Nicaragua, Costa Rica, Panama;
8. Ocotea austinii C. K. Allen, J. Arnold Arbor.
26: 350. 1945. TYPE: Costa Rica.
lustin Smith A 125 (holotype, F not seen; iso-
type, MO)).
Zarcero,
Ocotea irazuensis Lundell, Wrightia 5: 339. 1977. TYP
Costa Rica. Volcán Irazu, Proctor 32355 озы
LL not seen; isotype, MO!).
Best recognized by the raised reticulation on the
upper leaf surface, the inrolled leaf bases, which
are decurrent on the petioles, and the coriaceous,
frequently oblong leaves.
Distribution. Costa Rica and Panama, from
2000 to 3000 m.
9. Oc otea bajapazensis Lundell, Wrightia 6: 8.
1978. TYPE: Guatemala. Lundell & Contreras
— (holotype, LL
MO).
—
not seen; isotypes, F!,
This species can be recognized by its sparsely
appressed pubescent twigs, paniculate-cymose in-
florescences, and small flowers (ca. 5 mm diam.).
It is similar to O. heydeana, which has larger flow-
ers (6.5-9 mm diam.) and occurs at higher alti-
tudes.
Distribution. Guatemala, from 100 to 300 m.
Annals of the
Missouri Botanical Garden
10. Ocotea barbatula Lundell, Wrightia 5: 336.
1977. TYPE: Guatemala. Baja Verapaz: Lun-
dell & Contreras 19444 (holotype, LL not seen:
isotype, MO!).
This species is vegetatively similar to O. me-
ziana, but differs in having pit domatia with a pu-
bescent margin and well away from the midrib.
More collections may well show that these differ-
ences do not hold and that O. barbatula is better
treated as a synonym of O. meziana.
Distribution. Guatemala.
11. Ocotea bernoulliana Mez, Jahrb. Kónigl.
Bot. Gart. Berlin 5: 275. TYPE: Gua-
temala. Mujulia, Bernoulli & Cario 2590 (syn-
type, U!).
Best recognized by the following combination of
characters: twigs, leaves, terminal buds, and flowers
glabrous, leaves with conspicuous pit domatia in
the axils of lateral veins and smaller pit domatia
along the lateral veins, and cupules with 6 promi-
nent longitudinal ridges. This species is rarely col-
lected.
Distribution.
1000 to 1600 m.
Chiapas and Guatemala, from
12. Ocotea betazensis (Mez) van der Werff, No-
von 9: 572 39. Phoebe betazensis Mez,
Jahrb. Kónigl. Bot. Gart. Berlin 5: 192. 1889.
TYPE: Mexico. Oaxaca: Liebmann 2, 3, 22, 23
(syntypes, C!), Galeotti 2885 (syntype, not
seen), Jürgensen 575 (syntype, not seen).
Ore — mexicana Meissner var. diminuta Meissner,
Prodr. 15(1): 118. 1864. Ocotea mexicana
— r) Hemsley var. diminuta к Не m-
sley, Biol. Centr. Amer., Bot 1
Mexico. Jürgensen 575 (not see ш.
g
Characteristic are the rather broad, elliptic
leaves, relatively long (12-23 mm)
ments half as long as the anthers, and the densely
petioles, fila-
tomentellous or tomentose twigs. This species had
been included in O. helicterifolia, but that species
has frequently obovate leaves, shorter (4—15 mm)
petioles, sparse indument on the twigs, and nearly
sessile anthers.
Distribution. Oaxaca, to be expected in Chia-
pas, 2000—2600 m.
13. Ocotea botrantha Rohwer, Bot. Jahrb. Syst.
112(3): 375. 1991, non Ocotea matudai Lun-
de 2 Persea matudai Lundell, Lloydia 4: 49.
1941. TYPE: Mexico. Matuda 1880 (holotype,
E not seen; — MO».
Easily recognized by its racemose inflorescences,
arge flowers with tongue-shaped stamens, and
densely pubescent twigs. It is closely related to O.
sinuata, which differs in its paniculate-cymose in-
florescences and densely pubescent pistils.
Distribution. Chiapas, Guatemala, El Salvador,
from 800 to 2400 m.
14. Ocotea bourgeauviana (Mez) van der Werff,
Novon 9: 574. 1999. Phoebe bourgeauviana
Mez, ps Kónigl. Bot. Gart. Berlin 5: 194.
1889. TYPE: Mexico. Bourgeau
2234 Manos MO!).
Veracruz:
—
Wrightia 5: 34.
Nectandra — Lundell, )74.
Contreras 11186 (holo-
» PE: Guatemala. Izabal:
е, LL? 101 seen; isotype, MO!).
Phot aan Schultes, Bot. Mus. Leafl. 9: 170.
лппатотит chinantecorum (Se :hultes) Kos-
— Reinwardtia 6: 20. 1961. TYPE: Mexico.
Oaxaca: Schultes & Reko 827 (holotype, GH!).
The combination of clustered leaves and tepals
that are glabrous on the outside and pubescent on
the inside is diagnostic. It can resemble O. helic-
erifolia, but that species has the inner surface of
the tepals glabrous. The two collections from Gua-
temala and Honduras have the interior of the re-
ceptacle pubescent, while the Mexican specimens
have a glabrous interior of the receptacle. More col-
lections are needed to determine if the specimens
from Guatemala and Honduras are properly placed
in О, bourgeauviana.
Distribution. Mexico (Veracruz, Oaxaca, Chia-
pas). Belize, Guatemala, Honduras, from 100 to
1200 m.
15. — brenesii Standley, Publ. Field Mus.
at. . Bot. Ser. 18: 454. 1 . Nectandra
dedi "Gtandley) € C. K. Allen, 1. Arnold Ar-
bor. 26: 370. 1945. TYPE: Costa Rica. Brenes
13653 (holotype, F!).
The racemose inflorescences and appressed in-
dument on the young twigs characterize O. brenesit.
It is similar to O. pittieri (Mez) van der Werff, but
the latter species has a short, erect indument on
the twigs and domatia in the form of axillary tufts
of hairs.
Distribution,
2000 m.
Costa Rica, Panama, from 700 to
16. Ocotea calophylla Mez, Jahrb. Kónigl. Bot.
Gart. Berlin 5: 298. 1889. Pleurothyrium ve-
lutinum Meissner, DC. Prodr. 15(1): 170.
1864. TYPE: Colombia. Jervise s.n. (holotype,
K not seen).
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
439
ж — Standley & Steyermark, Ceiba 1: 237
TYPE: Costa Rica. Cartago: Leon 2166 (ho-
10!).
F!; isotype, MC
A striking species easily recognized by its dense-
ly pubescent leaves with a revolute base, the ver-
nation lines on the lower surface, and its upper
montane habitat.
Distribution.
from 2600 to 3000 m.
ica,
17. Ocotea cernua (Nees) Mez, Jahrb. Königl.
1889. ( „о пе
36. TYPE
Bot. Gart. Berlin 5: 377.
cernua Nees, Syst. Laurin, 424.
Martinique. Sieber 106 (syntype, MO!).
A common and widely distributed species with
nearly glabrous twigs, leaves, inflorescences, and
flowers;
nodding or reflexed. The cupules are deeply cup-
shaped:
veined.
Distribution.
Brazil, from O to 700 m.
18. Ocotea chiapensis (Lundell) Standley &
Stevermark. Publ.
ser. 23:
dell. Contr. Michigan Herb. 6: 12. 1941.
TYPE: Mexico. Chiapas: Matuda 22 (holo-
type, MICH not seen; isotype, MO!).
Field Mus. Nat. Hist.. Bot.
Univ.
Distinctive are the leaves with decurrent, in-
rolled bases, domatia as axillary tufts of hairs, and
the relatively large (5-6 mm diam.) flowers. It is
similar to O. glaucosericea Rohwer from Costa Rica
and Panama, but the latter species has a denser
indument on leaves and flowers and usually lacks
domatia.
Distribution. Mexico (Guerrero, Oaxaca, Chia-
pas). Guatemala, from 1000 to 2800 m.
19. Ocotea chrysobalanoides (Lundell) Lun-
dell. Wrightia 5: 35. Persea. chrysoba-
lanoides Lundell, Wrightia 1: 151. 1946.
TYPE: Mexico. Chiapas: Matuda 5582 (holo-
type. LL not seen: isotype, МО!).
Only known from the type collection. This spe-
cies can be recognized by its leaves with obtuse to
rounded apices, the pubescent inner surface of the
receptacle, large (0.8 mm long) staminodia, and the
dense, gray indument on twigs and inflorescences.
Distribution. Mexico (Chiapas). at 2000 m.
20. Ocotea congregata van der Werff, Novon 9:
514. 1999, TYPE: Mexico. Chiapas: Shilom
Ton 8930 (holotype. MO!).
Costa Rica. northern South Amer-
the unisexual flowers are often somewhat
leaves are frequently somewhat tripli-
From Mexico south to Bolivia and
. 1944. Nectandra chiapensis Lun-
Can be recognized by its clustered
leaves, the brown to dark brown indument on the
twigs, and the rather long (10 mm or more) petioles.
It can be confused with O. tonii (Lundell) van der
Werff, which differs in its shorter (to 6 mm) petioles
and yellowish brown indument on the young twigs.
Distribution. Mexico (Chiapas). 900 to
1400
loose ly
from
21. Ocotea contrerasii Lundell, Wrightia 5:
337. 1977. TYPE: Guatemala. Dept. Baja Ver-
apaz: Lundell & Contreras 19588 (holotype, LL
not seen: isotype, MO!).
A rarely collected species related to O. meziana.
from which it differs in the dense indument on the
twigs, with erect or ascending hairs and spreading
hairs along the major veins on the lower leaf sur-
face.
Distribution, Guatemala.
22. Ocotea eds i van der Werff. Novon 9:
TY
574. 199 PE: Mexico. Oaxaca: Wendt et
al. 6765 Prin] MEXU not seen; isotype,
MO).
Only known from the type collection and easily
recognized by its rugose-bullate. pubescent leaves.
lts racemose inflorescences are also a useful char-
acter.
Distribution.
1300 m
Mexico (Oaxaca), from 1100 to
23. Ocotea darcyi van der Werff, Novon 11: 505.
2001. TYPE: Panama. Correa & Montenegro
10176 (holotype, PMA not seen; isotype, MO!).
This species is closely related to O. stenoneura
Mez & Pittier, which differs in its inrolled and de-
current leaf bases, acute leaf apices, and elliptic
leaf shape. It can also be confused with O. pseu-
dopalmana Burger, but O. darcyi differs in having
many-flowered inflorescences, a ferruginous indu-
ment (not brown), and smaller flowers (4—4.5
vs. 6—8 mm diam.).
mm
Distribution. Panama, from 700 to 1000 m.
24. Ocotea eh es ras Mez, Jahrb. Kónigl.
Bot. Gart Berlin 5: 238.
macrophylla Beurl., ife
1854: 145. 1856, non Ocotea macro-
phylla HBK. TYPE: Panama. Billberg s.n. (ho-
lotype, 5 not seen).
389. Dendrodaphne
Vetensk. Acad.
Handl.
Ocotea ovandensis Lundell, Contr. Univ. Michigan Herb.
6: 16. 1941. TYPE: Mexico. Chiapas: Matuda 444
(holotype, MICH not seen).
440
Annals of the
Missouri Botanical Garden
Ocotea quisara Mez & Donnell Smith, Bot. Gaz. (Craw-
fordsville) 33: 259, 1902. ТҮРЕ: Costa Rica. J. Don-
nell Smith 6756 —— BM not seen), J. Donnell
Smith 6751 (syntypes, В, K not seen), J. Donnell
Smith 6753 (syntype, K not seen)
A widespread species, easily recognized by its
hollow twigs and tongue-shaped stamens with a
sterile tip. The poorly known O. morae Gomez-
Laurito shares these characters, but has leaves with
an inrolled base (flat in O. еш and
much larger fruits (5 X 3.5 em vs. 2.5 X ] cm in
O. dendrodaphne).
distribution. From Mexico to Panama, 50-
1300 m
25. Ocotea — van der Werff, Fieldiana,
Bot., n.s. 23: 79. 1990. TYPE: Costa Rica. Go-
mez & Herrera erie (holotype, MO”).
Distinctive are the obovate-elliptic leaves with
inrolled, decurrent bases and with a rather sparse,
erect indument on the lower surface, while the
twigs are appressed pubescent.
Distribution. Atlantic slopes of Nicaragua, Cos-
ta Rica, and Panama, from 100 to 1000 m.
26. Ocotea effusa (Meissner) Hemsley, Biol.
Centr. Amer, Bot. 3: 73. 1882. Oreodaphne
effusa Meissner, DC. Prodr. 15(1): 120. 1804.
TYPE: Mexico. Prope S. Pedro Nolasco, Talea
etc., Jürgensen 906 (syntype, BM!).
Best recognized by its lanceolate leaves with
shallow, often pubescent pit domatia. The large (0.8
mm) staminodia are also a useful character. Several
collections from higher altitude (1700-2000 m) are
provisionally placed here.
Distribution. Southern Mexico, Belize, Guate-
mala, from 200 to 2000 m.
27. Ocotea endresiana — Jahrb. Kónigl. Bot.
Gart. Berlin 5: 257. 1889. TYPE: Costa Rica.
Endres 223 — E,
Characterized by its glabrous leaves, inflores-
cences, and flowers, reflexed leaf bases, and cu-
pules without persistent tepals. Collections from
Costa Rica tend to have well-developed domatia,
but these are less conspicuous or even lacking in
specimens from Panama. Large-leafed specimens
& Williams, but that
species has puberulous flowers.
resemble O. rivularis Standl.
Distribution. Costa Rica, Panama, from 200 to
1500 m
28. Ocotea eucuneata Lundell, Contr. Univ.
Michigan Herb. 6: 16. 1941. TYPE: Belize.
Stann Creek Distr., Gentle 3068 (holotype,
MICH not seen; isotype, M
Only known from the type collection, Distin-
guishing features are the obovate leaves with a
sparse, erect indument on the lower surface, which
is best seen along the major veins, and the sharply
angled young twigs.
Distribution. Belize, at 200 m.
29. Ocotea euvenosa Lundell, Wrightia 4: 157.
1971. Ocotea venosa Lundell, Phytologia 12:
245. 1965, 1931. TYPE: Gua-
temala. Alta Verapaz: Contreras 4678 (holo-
non Gleason,
type, LL).
Known to me only from the type collection. Di-
agnostic are the large (to 27 X 15 cm) leaves, dry-
ing nearly black, with pinnate venation and pit
domatia in the axils of the basal lateral veins.
Closely related are O. bernouilliana, which has nar-
rower leaves and pit domatia not only in the axils
of the basal lateral veins but also along the lateral
veins, and O. vanderwerffii (Kostermans) van der
Werff, which has smaller, tripliveined leaves.
Distribution. | Guatemala.
30. Ocotea fendleri (Meissner) Rohwer, Mitt.
Inst. Allg. Bot. Hamburg 20: 152. 1986. Gym-
Prodr. 15(1):
Fendler 2395
nobalanus fendleri Meissner, DC.
142. 1864. TYPE: Venezuela.
(holotype. G-DC not seen).
Known in Mesoamerica from only two collec-
tions, a pistillate and a sterile one; among the spe-
cies with unisexual flowers best recognized by the
dense, shaggy pubescence on the twigs, the decur-
rent leaf bases, and the erect indument on the lower
leaf surface. Identifications of the two Mesoameri-
can collections are provisional.
French Guy-
Distribution. Panama, Venezuela,
ana, at 1000—1100 m.
31. Ocotea floribunda (Swartz) Mez, Jahrb.
Kónigl. Bot. Gart. Berlin 5: 325. 1889. Laurus
floribunda Sw., Prodr. Veg. Ind. Occ., 65.
1788. TYPE: Jamaica. Swartz s.n. (holotype, S
not seen).
A widespread species, best recognized by its pu-
bescent pistil or pistillode, the rather large (6—7
mm diam.) flowers, and the platelike cupule with a
double margin.
Distribution. From Nicaragua south to Brazil
and Peru, 100-1400 m.
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
32. Ocotea glaucosericea Rohwer, Mitt. Inst.
Allg. Bot. Hamburg 20: 144. 1986. Nectandra
hypoglauca Standl. ex С. К. Allen, J. Arnold
Arbor. 26: 399. 1945,
ees) Mez, 1889. TYPE: Panama. Chiriquí:
Davidson 531 (holotype, F not seen: isotype.
O!).
non O. hypoglauca
Similar to O. chiapensis: see that species for a
discussion of the differences.
Distribution. Costa Rica, Panama. from 1500 to
2000 m.
33. Ocotea ign Burger, Fieldiana, Bot., n.s..
23: 81. . TYPE: Costa Rica. Gomez- Jed
ito et al. pus (holotype, CR not seen: iso-
type, MO).
A very distinctive species with large (10—14 mm
diam.) flowers, densely pubescent twigs. leaves,
and flowers, and persistent tepals on the cupule.
Distribution. Costa Rica,
from 800 to 1400 m.
possibly Panama.
34. Ocotea gordonii van der Werff. Novon 9:
575. 1999, TYPE: Panama. Chiriquí: McPher-
son 10421 (holotype, MO”).
This species belongs to the O. helicterifolia group
and is characterized by its racemose inflorescences,
long (ca. 1 em) pedicels, and its pubescent inner
surface of the receptacle. Most other species in this
group have paniculate-cymose inflorescences. In
Werff
shares the racemose inflorescences, but that spe-
Panama only O. purpurea (Mez) van der
cies has smaller (5 mm vs. 8—10 mm diam.) flowers.
Distribution. Panama, from 1000 to 1150 m.
35. Ocotea guatemalensis Lundell, Wrightia 5:
339. 1977. TYPE: Guatemala. Baja Verapaz:
Lundell & Contreras 19754 (holotype. LL not
seen; isotype, MO!
—
A rarely collected species with rather large (to
25 X 8 cm) leaves.
raised reticulation on the lower leaf surface. The
conspicuous domatia, and
density of the pubescence on the inflorescences in-
creases markedly from the base toward the flowers.
A few collections from Costa Rica and Panama are
provisionally included here; they differ in having
smaller (ca. 3 mm vs. 4-5 mm diam.), less pubes-
cent flowers.
Distribution. Guatemala, possibly Costa Rica
and Panama.
36. Ocotea haberi van der Werff, Novon 11:
505. 2001. TYPE: Costa Rica. Guanacaste: 4—
5 km NW of Monteverde, Haber, Guindon &
Brenes 11093 (holotype, MO!).
An inconspicuous species best recognized by its
pubescent flowers, pit or pocket domatia, often with
a narrow opening, and a pubescent inner surface
of the receptacle. Most other species with pit dom-
atia have glabrous flowers and a glabrous inner sur-
face of the receptacle.
from 800 to
Distribution. Costa Rica, Panama.
1400 m
37. Ocotea hartshorniana Hammel, J. Arnold
Arbor. 67: 128. 1986. TYPE: Costa Rica. He-
redia: Hammel 11932 (holotype, DUKE not
seen; isotype, MO!).
Best recognized by the erect or ascending hairs
on the lower leaf surface, the dense, reddish brown
indument on twigs and inflorescences, and the de-
current leaf bases. It is very similar to О. montev-
which has 2
erdensis Burger.
2
predominantly ap-
pressed indument on the lower leaf surface апа
occurs at higher altitude. It can also be confused
with O. stenoneura, but that species has inrolled,
not flat leaf bases.
Distribution. Costa Rica, Panama, from 100 to
300 m.
38. Ocotea helicterifolia (Meissner) Hemsley,
Biol. Centr. Bot. 3: 73. 1882. Oreoda-
phne helicterifolia Meissner, DC. Prodr. 15(1):
123. 1864. Phoebe helicterifolia (Meissner)
Mez. Jahrb. Kónigl. Bot. Gart. Berlin 5: 193.
889. Cinnamomum helicterifolium (Meissner)
Kostermans, Reinwardtia 6: 21. 1961. TYPE
Mexico. Chiapas: Linden 1641 (syntype, K!).
Amer.,
Nectandra corzoana Lundell, Wrightia 4: 102. 1969. Phoe-
zoana (Lundell) Lundell, Wrightia 5: 342.
1977. ME corzoanum — Koster-
mans, Reinwardtia 10: 422. TYPE: Mexico.
Chiapas: Shilom Ton 3560 dine LL!
Ocotea tenejapensis Lundell, ightia 4: 108. 1969.
TYPE: Mexico. Chiapas: à Ton 779 (holotype.
d
x
x
I
LED.
Oreodaphne mexicana Meissner, DC.
1864
Prodr. 15(
Oreodaphne mexicana var.
sner, DC. Prodr. 15(1): 118. 1864. nom. supe
var. mexicana. Ocotea mexicana — Hemsl,
Biol. Centr. Amer., Bot. 3: 73. 1882. Ocotea mexi-
cana var. cy (Me занае) E ux as C entr.
Amer., Bot. 3: 73. 1882, nom. те rfl. -
сапа. — Rd euet Mez, Jahrb. Кош. ‘Be
Gart. Berlin 5: 194. 1889, non Phoebe mexicana
Meissner. TYPE: Mexico. Galeotti 7004 (lect
totype,
— here, BR!), Jiirgensen 537 (syntype, not
seen).
le LES.
subsessilis Meis-
3.
Annals of the
Missouri Botanical Garden
m mexicana Meissner и longipes Meissner,
DC. odr. 15(1): 118. 1864. Ocotea mexicana
коне r) Hemsley var. — ome — er) em-
sley, Biol. Centr. Amer., Bot. 3: 73. 1882. TYPE:
Mexico. Cerca Orizaba, Botteri 1018 | — K not
—
м
A
dins umbrosa Nees bullata Meissner, DC.
rodr. 1): 122. 1864. as umbrosa var. —
(Nees) Hemsley, Biol. Centr. Amer., Bot. 3: 74. 1882
me E: Central America. Oersted 21 ы В de-
royed).
Phoe "s — Lundell, Contr. Univ. Michigan Herb. 6:
21. 1941. Cinnamomum obtusatum 1 m Kos-
termans, Reinwardtia 6: 22. 1961. : Mexico.
Chiapas: Matuda 1887 (holotype. s not seen;
isotype, CAS!).
A widespread and variable species, character-
ized by the hirsute indument on twigs and leaves,
paniculate-cymose inflorescences, glabrous flowers,
alternate leaves, and a glabrous inner surface of the
receptacle. Collections placed here from Costa Rica
have obovate leaves and occur at lower altitudes;
they might represent a different species. Ocotea te-
nejapensis, only known from the type, 15 provision-
ally included here.
Distribution. From Mexico to Panama, at 1000—
1900 m (in Costa Rica at 50-600 m).
Lundellia 1: 40.
Wendt et al.
isotype,
39. Ocotea heribertoi Wendt,
1998. TYPE: Mexico. Oaxaca:
6871 (holotype, MEXU not
MO!)
seen,
Tall trees, known from only two collections and
best recognized by their smooth leaves with an ob-
tuse or rounded apex, presence of domatia at some
distance from the midrib, and its large fruits (to 4.5
cm diam.).
Distribution. Mexico (Oaxaca, to be expected in
Chiapas), at 200—300 m.
40. Ocotea heydeana (Mez & Donnell Smith)
Bernardi. Candollea 22: 93. 1967. Nectandra
heydeana Mez & Donnell Smith, Bot. Gaz.
(Crawfordsville) 19: 262. 1894. TYPE: Gua-
temala. Heyde & Lux 4260 (syntype, MO!).
Can be recognized by its rather large (6.5-9 mm
diam.), glabrous flowers, paniculate-cymose inflo-
rescences, and sparsely pubescent twigs. Similar to
О. bajapazensis, which has smaller (4.5-5.5 mm
diam.) flowers and occurs at lower altitudes, and to
O. magnifolia (Lundell) Lundell, which has a dens-
er indument on the twigs.
Hondu-
Distribution. Guatemala, El Salvador,
ras, from 600 to 1900 m.
41. Ocotea holdridgeiana Burger, Fieldiana,
Bot, n.s. 23: 83. 1990. TYPE: Costa Rica.
Lent 1677 (holotype, F not seen; isotype, MO!).
Easily recognized by the conspicuous pocket
domatia in the axils of the basal lateral veins, the
arge (10 mm diam.) flowers, and the stamens with
a sterile tip. Leaves can be somewhat tripliveined.
Distribution. Costa Rica, Panama, from 1600 to
2400 m.
42. Ocotea insularis (Meissner) Mez, Jahrb.
Kónigl. Bot. Gart. Berlin 5: 271. 1889. Phoebe
insularis Meissner, DC. Prodr. 15(1): 33. 1864.
TYPE: Costa Rica. Cocos Island, Menzies s.n.
(holotype, K not seen: isotype, MO!).
llen, J. pola Arbor. 26: 419.
TYPE: ia White 225 (holotype, МО).
п. ira Me & Pittier, Bull. Herb. 8:
232. . TYPE: Costa Rica. Tonduz in Herb. Inst.
Phys. i Costar. 10415 (syntype, US!), idem
13399 (syntype, US!).
Aiouea d C. K.
19
dér
Boiss.,
As accepted here, a variable species with rather
large leaves with decurrent bases, small (2.5-3.5
mm diam.) flowers with erect tepals, and tufts of
hairs on the back of the inner stamens. Montane
populations have often smaller, more coriaceous
leaves. Domatia are often, but certainly not always.
present.
Distribution. Costa Rica, Panama, Colombia,
and Ecuador, from 100 to 2000 m.
43. Ocotea iridescens Lorea-Hernandez & van
der Werff, Brittonia 54: 2002 in press. TYPE:
Mexico. Oaxaca: Salas.
Among the Mexican species of Ocotea, this spe-
cies is distinctive because of its dense, minute, and
light-colored indument on the lower leaf surface,
its acuminate leaves, and slender inflorescences. It
is known from a few collections made near the bor-
der of Oaxaca with Chiapas at 1400-1600 m.
44. Ocotea jefensis van der Werff, Novon 11:
506. 2001. TYPE: Panama. Prov. Panamá:
Cerro Jefe, Carrasquilla 2123 (holotype, MO!).
Diagnostic are the hollow twigs, obovate leaves,
and small (4 mm diam.), tomentellous flowers. Its
petioles are indistinct due to the decurrent leaf ba-
ses.
Distribution. Panama, from 200 to 900 m.
45. Ocotea jorge-escobarii Nelson, Ceiba 25:
173. 1984. TYPE: Honduras. Olancho: Nelson
& Soto 8188 (holotype. TEFH not seen; iso-
type, MO).
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
A montane species with dark green, stiff leaves,
pit domatia along the basal lateral veins, and large.
ribbed cupules. Closely related is O. pullifolia van
der Werff. which differs in leaves with obtuse or
rounded apices, and in having pit domatia restrict-
ed to the axils of the basal lateral veins.
Distribution. Honduras, Nicaragua, from 1000
to 1500 m.
46. Ocotea klepperae van der Werff. Novon 11:
508. 2001. TYPE: Costa Rica. Puntarenas:
Hammel 22068 (holotype, INB not seen; iso-
type, MO).
The large flowers, tongue-shaped stamens with a
sterile tip, and the position of the inflorescences in
axils of bracts near the tips of the branches place
this species in the subgenus Dendrodaphne. It dif-
fers from the other Mesoamerican species of this
subgenus in the dense indument on the twigs, in-
florescences, and flowers.
Distribution. Costa Rica (2 collections), from
10 to 300 m.
47. Ocotea laetevirens Standley & Steyermark,
Field Mus. Publ. Bot. 23: 114. 1944. TYPE:
Guatemala. Huehuetenango: Cerro Chiblac,
Steyermark 49378 (holotype, F!).
nov.
DS».
dias — Lundell, Wrightia 4: 133. 1970. Syn.
: Mexico. Chiapas: Clarke 38 (holotype.
As accepted here, a variable species that in-
cludes all specimens of the O. meziana group with
elabrous or sparsely appressed pubescent leaves,
pubescent terminal buds, and without domatia or
with domatia as axillary tufts of hairs. It is close to
О. meziana, but that species has pit domatia on the
lower leaf surface. Some collections from Costa
Rica and Panama that are vegetatively indistin-
guishable from O. laetevirens have 2-celled anthers:
in spite of this, they are here included in O. lae-
tevirens and not described as new species of Aiouea
Aubl.
Distribution. Mexico (Chiapas).
Costa Rica, and Panama, from 100 to
Guatemala,
Honduras,
3000 m.
48. Ocotea lentii Burger, Fieldiana, Bot., n.s. 23:
86. 1990. TYPE: Costa Rica. Cartago: Lent
/94 (holotype, F not seen; isotype, MO!).
Distinctive are the large (18—40 cm), obovate
leaves, sparsely pubescent leaves, receptacles pu-
bescent on the inner surface and the large (5 X 2.5
valerioides
ст) fruits. It can be confused with O.
Burger, but that species has the twigs and inflores-
cences completely covered by the indument, while
in O. lentii part of the surface remains visible. The
hairs of O. lentii are also coarser.
Distribution. Costa Rica, from 700 to 1400 m.
49. Ocotea leucoxylon (Swartz) Lanessan, Pl.
Util. Col. Franc. 156. 1886. Laurus leucoxylon
Sw.. Prodr. Veg. Ind. Occ. 65. 1788. TYPE:
Jamaica. Swartz s.n. (holotype, S not seen).
Ocotea — Lundell, Wrightia 5: 54. 1974. TYPE:
Guatemala. Izabal: Contreras 9924 (holotype. LL not
is — MO!).
— е — Publ. Field Mus. Nat. Hist
Bot. Ser. 18: TYPE: Costa Rica. fient
6789 — Ма
Diagnostic аге the small, unisexual flowers.
leaves with a smooth upper surface, and the mi-
nule, appressed, gray indument on the young twigs.
The flowers are less densely pubescent than the
pedicels. Cupules are shallow to platelike and often
lenticellate.
Distribution. From Mexico to Brazil and Peru,
West Indies, from 200 to 1600 m.
Ocotea macrantha van der Werff, Novon 11:
508. "РЕ: Costa Rica. Puntarenas:
Canton de Osa, Aguilar 4688 (holotype. MO!
50.
—
Vegetatively very similar to O. rubriflora Mez:
both have a short and dense indument on the twigs
and However, O.
arger flowers (10-12 mm vs. 6-7 mm diam.).
Costa Rica, from 100 to 200 m.
inflorescences. macrantha has
Distribution.
51. Ocotea macropoda (HBK) Mez. Jahrb. Kón-
igl. Bot. Gart. Berlin 5: 348. 1889. Persea ma-
cropoda HBK, Nov. Gen. Sp. 2: 127. 1817.
TYPE: Colombia. Humboldt & Bonpland s.n.
holas P not seen).
Mem. New York Bot. Gard.
TYPE: Venezuela. Bernardi
Ocotea babosa С. Allen,
9: 82. 1966. pis nov.
86 (holotype, NY!
Best recognized by the unisexual flowers and
erect indument on twigs, leaves, and flowers. In
these characters O. macropoda resembles O. atlan-
tica, but in the latter species the indument covers
the surface of the twigs completely, while it remains
partly visible on O. macropoda. Ocotea macropoda
also has larger flowers than O. atlantica (5-6 mm
vs. 3.5 mm diam.). I am not quite certain that the
name O. macropoda is correctly applied here: the
South American specimens (including the types of
O. macropoda and О. babosa) come from much
higher altitude (above 2000 m).
444 Annals of the
Missouri Botanical Garden
Distribution. Costa Rica, Panama, Colombia, pubescence, and the densely pubescent twigs. It is
Venezuela, Ecuador, from 100 to 800 m. infrequently collected.
Distribution. Costa Rica, from 1400 to 2300 m.
52. Ocotea magnifolia (Lundell) Lundell,
Wrightia 5: 341. 1977. Nectandra magnifolia
E: Gua-
Lundell, Wrightia 4: 103. 1969. TYP
temala. Alta Verapaz: Contreras 7865 (holo-
type, LL not seen; isotype, MO").
Nectandra thornei Lundell, Wrightia 5: 335. 1977. Syn.
TYPE: Mexico. Chiapas:
Thorne & Lathrop
40526 (holotype, LL not seen; isotype, DS!).
A poorly known and poorly defined species, best
recognized by its indument, which is denser than
in O. heydeana and sparser than in O. rubriflora
Mez.
Distribution. Mexico (Chiapas), Guatemala,
from 100 to 200 m.
53. Ocotea matudai Lundell, Bull. Torrey Bot.
e 69: 388. 1942. TYPE: Mexico. Chiapas:
. Ovando, Matuda 4221 (holotype, MICH
not seen; isotype, F!).
A rarely collected species best recognized by its
glabrous terminal buds and leaves; flowers have
spreading tepals at anthesis. The distal parts of the
inflorescences have a minute, predominantly erect
indument, also a useful character.
Distribution. Mexico (Chiapas), 1300-2700 m.
54. Ocotea УРИ К. Allen, J. Arnold Ar-
bor. 26: 1945. TYPE: Costa Rica. Zar-
cero, pines Smith H359 (holotype, F not seen;
isotype, MO!).
A rather common species, characterized by its
greenish drying leaves with small pit domatia along
the lateral veins, pubescent terminal buds, and gla-
brous flowers. It is similar to O. viridiflora Lundell,
which differs in its glabrous terminal buds, finely
acute leaf apices, and smaller inflorescences (to 5
em in O, viridiflora, to 10 em in O. meziana). Also
similar to O. laetevirens, but this species lacks pit
domatia
Distribution. Honduras, Nicaragua, Costa Rica,
and Panama, from 100 to 1800 m.
55. Ocotea mollicella (Blake) van der Werff,
Fieldiana, Bot., n.s. 23: 88. 1990. Phoebe mol-
licella Blake, Contr. Gray Herb. 52: 64. 1917
Cinnamomum mollicellum (Blake) Koster-
mans, Reinwardtia 6: 22. 1961. TYPE: Costa
Rica. Copey, Tonduz 11676 (holotype, GH!).
Readily recognized by its racemose inflorescenc-
es, elliptic to narrowly elliptic leaves with soft, gray
56. Ocotea — Mez & а Bull. Herb.
Boissi ser. 2, 3: 233. 19( TYPE: Costa
Rica. om A Er " not seen).
Best recognized by its densely pubescent twigs,
pubescent, chartaceous leaves, and many-flowered
inflorescences. See also under O. darcyi and O.
pseudopalmana Burger.
Costa Rica, Panama, 50 to
istribution. from
1000 m.
57. Ocotea ЕТЕ Burger, Fieldiana,
ot., n.s. 23: 89. 1990. TYPE: Costa Rica.
Puntarenas: Pietas dd (holotype, CR not
seen; isotype, MO!)
Similar to O. hartshorniana; see that species for
differences.
Distribution. Costa Rica, from 800 to 1500 m.
Novon 7:
Costa Rica. Alajuela: Go-
USJ not
58. Ocotea morae Gomez-Laurito,
145. 1997. TYPE:
mez-Laurito & Mora 12817 (holotype,
MO!).
seen; Isotype,
The tongue-shaped stamens and double-mar-
gined cupule place this species in subgenus Den-
drodaphne. lt can be confused with the widespread
О. dendrodaphne, but differs Е its inrolled leaf
base and large (5 X 3.5 cm) fru
Distribution. Costa Rica, a 100 to 800 m.
59. Ocotea multiflora van der Werff, Novon 6:
481. 1996. TYPE: Costa Rica. Puntarenas:
Reserva Forestal Golfo Dulce, Aguilar 791
(holotype, MO”).
Tall trees, best recognized by its leaves with
many (12 to 17) pairs of lateral veins, the many-
flowered (2-2.5 mm
diam.) flowers. It was included in Burger and van
der Werff (1990) as Ocotea sp. B.
Distribution. Only known from the Osa Penin-
sula in Costa Rica, from 10 to 200 m.
inflorescences, and small
60. Ocotea nigrita (Lundell) Lundell, Wrightia
5: 341. 1977. Nectandra nigrita Lundell,
Wrightia 4: 132. 1970. TYPE: Guatemala. El
Petén: Contreras 9465 (holotype, LL not seen;
isotype, CAS!)
An infrequently collected species, characterized
by its oblanceolate, black-drying leaves with tufts
Volume 89, Number 3
2
van der Werff
Synopsis of Ocotea
of hairs in the axils of the basal lateral veins and
the small (3—4 mm diam.), densely puberulous flow-
ers. This
apana Wendt & van der Werff, which differs in its
5-2.5 em in O.
s species can be confused with O. uxpan-
longer and wider leaves (6-15 X 1.
nigrita, 10-25 X 2.58 ст in 0. uxpanapana) and
larger (3—4 mm vs. 4-6 mm diam.) flowers.
Distribution. Guatemala, from 200 to 300 m.
61. Ocotea oblonga (Meissner) Mez, Jahrb. Kón-
igl. Bot. Gart. Berlin 5: 367. 1889. Mespilo-
daphne oblonga Meissner, DC. Prodr. 15(1):
107. 1864. TYPE: French Guyana, Sagot 491
(holotype, G-DC not seen).
Phoebe mayana Lundell, Amer. Midl. Naturalist 20: 473.
1943. Ocotea mayana (Lundell) Pond П, Wrightia 2:
52. 1960.
Reinwardtia 10: 446. 1988.
— по!
Cinnamomum der igi (Lundell) Kos-
TYPE: Belize.
seen: isotype,
termans,
Gentle 3187 (holotype,
10!).
This species can be easily recognized by the
combination of unisexual flowers and leaves with
pit or slit domatia. The slender, appressed pubes-
cent vegetative buds are also a good character.
Distribution. From Mexico to Bolivia and Bra-
zil. from O to 1000 m.
62. Ocotea oblongifolia van der Werff, Novon
11: 509. 2001. TYPE: Guatemala. Quezalte-
nango: Finca St. John, ca. 5 km S of Sta. Maria
de Jesus. Walker 442 dela US!
—
Diagnostic are the oblong, densely pubescent
leaves with an obtuse to subcordate base and the
small (2-3 mm diam.) flowers. The indument is
similar to. what in the O. helicterifolia
group, but the small flowers with erect to half-erect
is found
tepals are entirely different.
Only known from the type collec-
1400-1800 m.
Distribution.
tion, Guatemala,
63. Ocotea parvula (Lundell) van der Werff. No-
von 11: 510. 2001. Phoebe parvula Lundell,
Wrightia 5: 343. 1977. Cinnamomum parvul-
um (Lundell) Kostermans. Reinwardtia 10:
448. 1988. TYPE: Mexico. Chiapas: Ton 605
(holotype, LL).
An infrequently collected species with ovate
leaves, gradually tapering into the tip, domatia as
axillary tufts of hairs, glabrous tepals, and small (to
5 cm long) inflorescences with frequently persisting
bracts. See also O. strigosa van der Werff.
Mexico (Chiapas, Oaxaca),
Distribution. from
1800 to 2800 m.
64. — patula van der Werff, Novon 9: 577.
: Costa Rica. Puntarenas: Aguilar
et e. 2715 Delos: MO!
Part of the O. helicterifolia group and similar to
). valeriana (Standl.) Burger, from which it differs
in its densely tomentellous inflorescences (with the
surface almost entirely or entirely covered), its
shorter (2-3 mm vs. 6-8 mm) pedicels, and its pu-
bescent flowers.
Distribution. Known from two collections, Cos-
ta Rica, 1000-1400 m.
65. Ocotea pausiaca Rohwer, Bot. Jahrb. Syst.
112: 387. 1991. TYPE: Panama. Colón: Knapp
5782 (holotype, MO!
—
Part of the O. heydeana group, characterized by
rather large, glabrous or nearly glabrous flowers
with spreading tepals, these with some papillae on
the inner surface, and often dark-drying leaves. It
is distinct within this group by its stamens with the
filaments twice as long as the anthers and the outer
six stamens curved inward. like
O. rubriflora, which differs in the very dense, mi-
nute indument on the young twigs.
Distribution. Panama, from 900 to 1500 m.
Vegetatively it is
66. Ocotea сог »rum Gomez-Lauri-
. Novon 3: 31 1993. TYPE: Costa Rica. San
— E et al. 12160 (holotype, CR
not seen: isotype, МО!).
A very distinctive species characterized by its
stamens with a sterile tip and the densely tomen-
tellous whitish lower leaf surfaces. It was included
in Burger and van der Werff (1990) as “A species
f uncertain. generic position” and in Rohwer
(1991) as Ocotea sp. /
Distribution. Costa Rica, Panama, from 1800 to
2200 m
>
67. Ocotea pittieri Me z) van der Werff, Fieldi-
Bot.. n.s. 23: 92, 1990. Phoebe pittieri
. Bot. Jahrb. Syst. 7 Beibl. 67: 16. 1901.
bum pittieri (Mez) Kostermans, Rein-
wardtia 6: 23. 1961. TYPE: Costa Rica. Ala-
juela: Tonduz 11893 (holotype, B not seen).
Best recognized by its racemose inflorescences
and the dense, short, erect indument on the young
twigs. The indument on the lower leaf surface can
e very sparse and difficult to see. Ocotea pittieri
—
is very similar to O. brenesii, which has appressed
hairs on the twigs. The name 0. pittieri was applied
in Burger and van der Werff (1990) to the species
here treated as O. praetermissa van der Werff.
446 Annals of the
Missouri Botanical Garden
Distribution. Costa Rica. from 1800 to 2200 m. Ocotea ra — ех Brandegee, Univ. Calif. Publ.
68. Ocotea platyphylla (Lundell) Rohwer, Bot.
Jahrb. Syst. 112: 390. 1991. Phoebe platy-
P Lundell, Contr. Univ. Michigan Herb.
ii 23. 1941. )
Nectandra dU (L pies
. K. Allen, J. Arnold Arbor. )2. 1945.
m Mexico. Chiapas: Villes 1930 ciii
type, MICH not seen; isotype, MO!).
Very similar to O. heydeana, from which it differs
in its narrowly inrolled or revolute leaf bases.
Distribution. Mexico (Chiapas), Guatemala, El
Salvador, from 900 to 2600 m.
69. Ocotea praetermissa van der Werff, Novon
6: 482. 1996. TYPE: Costa Rica. Cartago:
Burger et al. 12065 (holotype, MO!).
Similar to O. purpurea, but O. praetermissa can
be recognized by its paniculate-cymose (not race-
mose) inflorescences and its glabrous (not pubes-
cent) flowers.
Distribution.
3200 m
Costa Rica, Panama, from 2000 to
70. Ocotea producta (C. K. Allen) Rohwer, Mitt.
Inst. Allg. Bot. Vie 20: 143. 1986. Nec-
tandra — ta С. К. Allen, J. Arnold Arbor.
26: 352. 1945. TYPE: dosis Rica. Prov. San
José: Skutch 3906 (holotype, GH not seen; iso-
type.
Only known from the type collection and char-
acterized by the long (4—5 ст) apparent petioles
with inrolled, decurrent leaf bases. The filaments
of the inner 3 stamens are fused.
Distribution, Costa Rica, 700 m.
71. Ocotea с Burger, Fieldiana,
Bot., n.s. 23: 92. . TYPE: Costa Rica. San
José: Lent 1679 — ED.
Diagnostic are the coriaceous leaves with erect
indument on the lower surface, the few-flowered in-
florescences, the large (6-8 mm diam.) flowers. and
confused with
the montane habitat. It can be
mollifolia, which has chartaceous leaves, smaller
(3-6 mm diam.) flowers, and occurs usually below
1000 m.
Distribution. Costa Rica, Panama, from 2200 to
m.
72. Ocotea puberula (Richard) Nees,
Laurin. 472. 1836. Laurus puberula Richard,
Actes Soc. Hist. Nat. Paris 1: 108. 1792.
TYPE: French Guyana. Le Blond. s.n. (holo-
)
Syst.
—
type, P not seen).
Bot. 7: 326. 1920. TYPE: Mexico. Veracruz: Purpus
8456 Ae i S not seen; isotype, MO!).
A widespread species, best recognized by its
unisexual flowers, the slightly raised reticulation on
the upper leaf surface, and the rather large, (thinly)
chartaceous leaves. Ocotea leucoxylon has the up-
per surface of the leaves generally smooth and
smaller (3—4 vs. 4—5 mm diam.
Distribution. From Mexico to Argentina (in Me-
soamerica known from Chiapas, Costa Rica, and
flowers.
—
Panama), from 0 to 1300 m.
73. Mentes pullifolia van der Werff, Novon 11:
509. 2001. TYPE: Costa Rica. Puntarenas:
Canto de Golfito, Herrera 4119 (holotype,
INB).
Closely related to O. jorge-escobarii from Hon-
duras and Nicaragua. Ocotea pullifolia is best rec-
ognized by its obtuse or rounded leaf apices and in
having pit domatia restricted to the axils of the low-
ermost lateral veins. In some Panamanian collec-
tions the pit domatia are lacking.
Distribution. Costa Rica, Panama, from 200 to
1100 m
74. Ocotea purpurea (Mez) van der Werff, No-
von 9: 579. 1999. Phoebe purpurea Mez, Jahrb.
Königl. Bot. Gart. Berlin 5: 196. 1889. TYPE:
Guatemala. Alta Verapaz: von Tuerckheim 371
(syntypes, B, K, LE not seen, photo MO!).
Nectandra о Lundell, Wrightia 5: 33. 1974.
TYPE: Guatemala. Contreras 11235 (holotype, LL
not seen; (gem ).
Readily identified by its usually racemose inflo-
rescences, rather small (to 11 em long) leaves, and
sparsely pubescent flowers. Specimens from Pana-
ma differ in having fewer lateral veins (mostly 4
pairs
Distribution. Mexico (Oaxaca, Chiapas), Gua-
temala, Honduras, Panama, from 1400 to 2600 т.
75. Ocotea racemiflora Lundell, Wrightia 4:
107. 1969. TYPE: Guatemala. Alta Verapaz:
Contreras 7904 (holotype, LL!).
Only known from the type collection, which is in
fruit with a few detached, old flowers. The flat, thin
cupules with persistent tepals and roundish fruits
are not known from other Mesoamerican species.
Distribution. Guatemala.
76. Ocotea rhytidotricha Rohwer, Bot. Jahrb.
Syst. 112: 391. 1991. TYPE: Nicaragua. Ma-
tagalpa: Hall & Bockus 7919 (holotype, MO!).
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
Part of the O. heydeana group and characterized
by the short, erect indument on the lower leaf sur-
face. The hairs are inconspicuous, not discernable
to the touch, and are best seen along the major
veins.
Distribution. El Salvador, Honduras, Nicara-
gua, from 1200 to 1600 m.
77. Ocotea rivularis Standley & L. O. Williams.
Ceiba 1: 238. 1951. TYPE: Costa Rica. Pun-
tarenas: Allen 5590 (holotype, EAP not seen:
isotype, MO!).
Readily recognized by its large (30—50 cm long).
obovate leaves and pubescent flowers. It can be
confused with O. endresiana, which has smaller
(10-30 em long) leaves and glabrous flowers.
Distribution. Costa Rica, from 50 to 300 m.
78. Ocotea rovirosae Lorea-Hernandez & van
der Werff, Brittonia 54: 2002 in press. TYPE:
Mexico. Veracruz.
This species is best recognized by the combi-
nation of its short, erect indument on twigs and
leaves, the paniculate-cymose, many-flowered inflo-
4—35
Nectandra lundellii
rescences, and the rather large ( em long),
re green drying leaves.
. К. Allen.
n large. yellow-green drying leaves, but has typ-
known from the same area, has simi-
ical Nectandra flowers with the inner surface of the
tepals and the stamens papillose: its tepals are also
spreading (not erect or half-erect) at anthesis.
Vera-
Distribution. Mexico (Oaxaca, Tabasco.
cruz), from 10 to 200 m.
79. Ocotea rubriflora Mez, Jahrb. Kónigl. Bot.
Gart. Pre 5: 279. 1889. Nectandra rubriflora
. K. Allen, J. Arnold Arbor. 26: 372
—
(syntype, G!).
Mexico. Tabasco: — s.Hn.
Ocotea perseifolia Mez & ge Smith, Bot. Gaz. (Craw-
fordsville) 20: 1¢ ҮРЕ: Guatemala. Izabal:
J. Donnell Smith 1807 piedad B not seen, US not
seen).
This belongs to the O. heydeana group in which
it stands apart by the dense and minute indument
on the twigs and inflorescences, which covers the
close to O.
crantha, but that species has larger flowers (10—12
surface completely. Vegetatively ma-
—
vs. 6—7 mm diam.
Distribution. From southern Mexico to Panama,
at 100-350 m
80. Ocotea rubrinervis Mez, Jahrb. Kónigl. Bot.
Gart. Berlin 5: 351. 1889. TYPE: Peru. Spruce
4580 (syntype, B!); Panama. Duchassaing s.n.
(syntype, not seen).
known from the Pacific
—
n Mesoamerica only
coast of Panama and best recognized by its unisex-
ual flowers, the readily visible gland dots on the
upper leaf surface, and the shallow cupules with
persistent tepals.
Distribution. Panama, from 0 to 50 m.
Ocotea rufescens van der Werff, Novon 6:
479. 1996. TYPE: Costa Rica. Limón: Aguilar
& Schmidt 1077 (holotype, MO!).
81.
The only Ocotea species in Costa Rica and Pan-
ama (where it is expected to grow as well) with a
dense, reddish brown or reddish indument on twigs
and lower leaf surface. Ocotea salvinii Mez, from
Mexico and Guatemala, has a similar indument but
has inrolled leaf bases.
Distribution. Costa Rica, from 450 to 1500 m.
82. Ocotea salvadorensis (Lundell) van der
Werff, Novon 6: 481. 1996.
dorensis Lundell, Wrightia 4:
Vectandra salva-
105. 1969. Phoe-
be salvadorensis (Lundell) Lundell, Wrightia 5:
344. TYPE: El Allen
7173 (holotype, LL not seen: isotype, GH!).
Salvador. Р.
Only known from the type collection. The dense,
tomentellous, gray-brown indument on the lower
leaf surface is diagnostic. Such an indument also
occurs in O. pharomachosorum, which has tongue-
shaped stamens with a sterile tip. and O. iridescens,
which has domatia. A few collections from montane
cloud forest in Honduras will key to O. salrador-
ensis, but differ from that species in their reddish
brown indument on twigs and leaves, and the rather
dense indument on the inner surface of the tepals.
More collections are needed to determine if the
specimens from Honduras represent an unde-
scribed specie
Distribution. El Salvador, at 2500 m.
83. Ocotea salvinii Mez. Jahrb. Kónigl. Bot.
Gart. Berlin 5: 264. 1889. Phoebe. salvinii
(Mez) Lundell, Contr. Univ. Michigan Herb. 6:
23. 1941. Cinnamomum salvinii (Mez) Koster-
Reinwardtia 6: 23. 1961. TYPE:
temala. Salvin s.n. (syntvpes, K not seen, W
mans, Gua-
[1
/
not seen).
Diagnostic are the ferruginous tomentellous
leaves with an inrolled base and the perfect flowers.
See also under O. rufescens.
448
Annals of the
Missouri Botanical Garden
Distribution, Chiapas and Guatemala, from
2400 to 3000 m.
84. Ocotea sinuata (Mez) Rohwer, Bot. Jahrb.
Syst. 112(3): 373. 1991. Nectandra sinuata
Mez, Jahrb. Kónigl. Bot. Gart. Berlin 5: 402
(1889). TYPE: Guatemala. Bernoulli & Cario
2581 (syntypes, B, GOET, K, LE all not seen).
Best recognized by its large (13-16 mm diam.)
flowers, anthers with a sterile tip and densely
brown-tomentose twigs. It can be confused with O.
botrantha; see that species for differences.
distribution. From southern Mexico to Panama,
but not reported from Honduras and Nicaragua,
from 200 to 1500 m.
85. Ocotea standleyi C. K. Allen, J. Arnold Ar-
bor. 26: 343. 1945. Phoebe — nan
dley & Steyermark, Field M "ubl. Bot.
116. 1944, non Phoebe mac dE (Blume)
Blume. 1851. TYPE: Guatemala. Alta Vera-
paz: Standley 70009 (holotype, F not seen).
Readily recognized by its large (15-35 cm long
М
leaves with a rounded to subcordate base, апа the
glabrous terminal buds and leaves. A Licaria spe-
cies occurring in Chiapas is vegetatively similar,
but its flowers (with only 3 stamens) and cupule
(deeply cup-shaped with a weakly developed dou-
ble margin) set it apart. The Licaria species gen-
erally occurs below 1000 m.
Distribution. Mexico (Chiapas), Guatemala,
from 800 to 2500 m, mostly above 1500 m
86. Ocotea stenoneura Mez & Pittier, Bull.
Herb. 2, 3: 233. 1903. TYPE:
Costa Rica. San Jose de Dota, Tonduz 13377
(lectotype, designated by Allen (1945: 334),
GH not seen; isolectotype, US!).
Boissier, ser.
A rarely collected species with decurrent, revo-
lute leaf bases, an erect, rather dense indument on
the lower leaf surface, and prominently raised ve-
nation on the lower leaf surface. Specimens placed
here vary in cupule shape, and more collections are
needed to determine if two species are involved.
Distribution. Costa Rica, Panama, possibly Co-
lombia and Ecuador, from 700 to 1700 m.
87. Ocotea strigosa van der Werff, Ann. Missou-
ri Bot. Gard. 75: 723. 1988. TYPE: Nicaragua.
Matagalpa: Stevens 22181 (holotype, MO!
—
An inconspicuous species with ovate leaves and
few-flowered inflorescences. It is similar to O. par-
vula, but that species has domatia and glabrous (not
sparsely pubescent) flowers. Ocotea iridescens is
also a close relative, but differs in its dense indu-
ment on the lower leaf surface
Distribution. Nicaragua, from 1000 to 1600 m.
88. Ocotea subalata Lundell, Lloydia 4: 48.
1941. TYPE: Mexico. Chiapas: Matuda 2957
(holotype, MICH not seen; isotype, F!).
Useful char-
acters for identification are the long (to 20 cm) in-
florescences and the sharply angled or winged,
Only known from two collections.
densely pubescent young twigs.
Distribution. Mexico (Chiapas), from 2100 to
2500
89. Ocotea tenera Mez & Donnell Smith, Bull.
Herb. Boissier, ser. 2, 3: 234. 1903. TYPE:
Costa Rica. Pittier 13396 (syntype, US!).
Oc ba — Lundell, Wrightia 6: 9. 1978. Syn. nov.
Guatemala. Lundell & Contreras 20948 (ho-
ee LL not seen; isotype, !
Ocotea eucymosa Lundell, Wrightia 5: : 35. 1974. Syn. nov.
TYPE: Guatemala. Contreras 11215 (holotype, LL
not seen; isotype, MO!
This species is best recognized by its small size
(to 12 m tall), dark-drying leaves with gland dots
on the upper leaf surface, glabrous terminal buds,
and small, glabrous flowers with erect or half-erect
tepals.
Distribution. Guatemala and Costa Rica, from
100 to 1600 m.
90. Ocotea tonduzii Standley, Field Mus. Publ.
Bot. 18: 456. 1937. Ocotea cuneata Mez, Bot.
Jahrb. 30, Beiblatt 67: 17. 1901, non Ocotea
cuneata (Grisebach) Gomez. 1889. TYPE:
Costa Rica. Tonduz 2142 (syntype, BR!).
reddish indument
on the flowers, the large, glabrous (or finely ap-
Characteristic are the dense,
pressed pubescent) terminal buds, the raised ter-
tiary venation on the lower leaf surface, and the
glabrous, nearly sessile leaves with an inrolled
base. Ocotea endresiana has a similar appearance
but lacks the reddish indument of the flowers.
Distribution. Costa Rica, from 1500 to 2400 m.
91. Ocotea tonii (Lundell) van der Werff, Novon
9: 579. 1999. Nectandra tonii Lundell, Wrigh-
tia 4: 106. 1969. TYPE: Mexico. Chiapas: Shi-
lom Ton 2014 (holotype, LL not seen; isotype,
NY!).
Distinctive are the clustered leaves with short (to
6 mm long) petioles, yellow-brown indument, pa-
Volume 89, Number 3
2002
van der Werff
Synopsis of Ocotea
niculate-cymose inflorescences, and pubescent
flowers. See also under O. congregata.
Distribution. Mexico (Chiapas). from 1000 to
1500 m.
92. Ocotea truncata Lundell, Phytologia 12:
244. 1965. TYPE: Guatemala. Alta Verapaz:
bu "731 (holotype, LL not seen; iso-
type, US!
м
This species сап be recognized by its thin, dark
green drying leaves with rather conspicuous dom-
atia and its few-flowered inflorescences with gla-
brous flowers. The truncate apex of the young fruits
is less pronounced in mature fruits and scarcely
helps in the identification of this species.
Distribution. Mexico (Chiapas),
from 700 to 1200 m.
Guatemala,
93. Ocotea uxpanapana Wendt & van der Werff,
nn. Bot. Gard. 74: 413. 1987.
TYPE: Mexico. Veracruz: Mpio. Minatitlan,
Wendt et al. 2869 (holotype, MEXU not seen:
isotype, MO!).
Missouri
Currently only known from the Uxpanapa region
in Veracruz. It resembles O. eucuneata and O. ni-
grita but differs in its rather large (to 25 em long).
elliptic larger. flowers. The
strongly lobed cupules appear to be a unique fea-
ture of this species, but cupules of O. eucuneata
and O, nigrita are not yet known.
Distribution. Mexico (Veracruz), to be expected
in Chiapas, from 100 to 300 m.
leaves and slightly
94. Ocotea valeriana (Standley) Burger, Fieldi-
ana, Bot., n.s. 23: 96. 1990. Phoebe valeriana
Standley, Publ. Field Mus. Nat. Hist., Bot. Ser.
18: 460. 1937. Cinnamomum valerianum
(Standley) Kostermans, Reinwardtia 6: 24.
1961. TYPE: Costa Rica. Tonduz 11746 (ho-
lotype, F not seen
м
— austinit €. K. Allen, J. Arnold Arbor. 26: 374.
n TYPE: Costa Rica. Alajuela: Austin Smith P
2 — A not seen).
— smithii C. Allen Arnold Arbor. 26: 317.
f an niidina Kostermans, Rein-
— 6: 2; . 1961, non Cinnamomum smithii Lu
manoff. : Costa Rica. Alajuela: Austin Smith
P.C. 307 Чаки F not seen).
Similar to O. helicterifolia, from which it differs
in its indument of the twigs (short matted hairs and
longer, erect ones in O. valeriana vs. only long,
erect ones in O. helicterifolia), its deeper cupules,
and its presence at higher altitudes.
Distribution.
to 2200 m.
Costa Rica and Panama, from 800
95. Ocotea valerioides Burger, Fieldiana, Bot..
n.s. 23: 97. 1990. TYPE: Costa Rica. Harts-
horn 1530 (holotype, СК not seen; isotype,
MO!
Distinctive are the large (to 40 em long) leaves
and the dense, tomentellous indument on the twigs
and inflorescences. Differences with O. lentii are
discussed under that species.
Distribution.
500 m.
Costa Rica, Panama, from 50 to
96. Ocotea vanderwerffii (Kostermans) van der
Werff, Novon 11: 511. 2001. Phoebe —
van der Werff. Ann. Missouri Bot. Gard.
406. 1987, non Ocotea glabra van der on
Cinnamomum vanderwerffit Kostermans, Rein-
wardtia 10: 454. 1988, non Cinnamomum gla-
brum Ettinghausen. TYPE:
Wendt et al. 4813 (holotype, N
exico. Oaxaca:
MO!).
Ocotea vanderwerffii is best recognized by its
dark-drying, tripliveined leaves with large pit dom-
atia in the axils of the basal lateral veins and the
glabrous flowers with erect tepals. It resembles 0.
tenera, but that species has pinnately veined leaves
and small pit domatia, which are not visible on the
upper surface
Mexico
Distribution. Southern
1200 m.
from 150 to
97. Ocotea br iar (Meissner) Mez, Jahrb.
Kónigl. Bot. Berlin 5: 240. 1889. Sas-
safridium veraguensis Meissner, DC.
15(1): 171. 1864. TYPE: Costa Rica. Bridges
s.n. (syntype, K not seen), Oersted, Laur. 10
(syntype, В not seen), Warszewicz 1 (syntype,
G not seen).
Gart.
Prodr.
65. 1917.
Ocotea bakeri Blake, Contr. Gray
РЕ GH not
TYPE: Nicaragua. ` Boker 2493 eh
een).
Ocotea a rines L unde Il, Contr. Univ. Michigan Herb.
6: 941. TYPE: Mex Chiapas: Matuda 654
(h — MICH not see
Ocotea paradoxa Mez, Bot. Jahrh. 30. Beibl. 67: 16. 1901.
TYPE: Tonduz in Herb. Phys.-
Geogr. Costar. 7648 (holotype, BR).
xico.
Costa Rica.
A widespread and rather frequently collected
species, best recognized by its papillose, tongue-
shaped stamens with a sterile tip, solid stems, gla-
brous (or nearly so) leaves, and its cupules with a
double
margin. Can be confused with O. dendro-
450
Annals of the
Missouri Botanical Garden
daphne, which has hollow twigs and a pubescent
inside of the receptacle.
Distribution. From southern Mexico to Panama,
from 10 to 2000 m.
98. Ocotea verapazensis Standley & Steyer-
Field Mus. Publ. Bot. 23: 114. 1944.
ҮРК: Guatemala. Standley 71421 (holotype,
En
A poorly known species with glabrous terminal
buds, twigs, and leaves, and slightly winged peti-
Vegetatively similar to О, subalata, but that
species has densely pubescent terminal buds and
oles.
pubescent inflorescences and flowers.
Distribution. | Guatemala, Eon 1500 to 2000 m.
99. Ocotea verticillata Rohwer, Bot. Jahrb. Syst.
112(3): 369. 1991. TYPE: Mexico. Veracruz:
Ibarra 2328 (holotype. MO!).
Characteristic are the whorled leaves abruptly
rounded at the base and its tongue-shaped stamens
with a sterile tip. Related to O. botrantha and O.
sinuata, both of which have alternate leaves.
Distribution, Mexico (Veracruz, Oaxaca), from
100 to 1000 m.
100. Ocotea viridiflora Lundell, Wrightia 5: 36.
1974. TYPE: Panama. Chiriquí: Proctor 31916
(holotype, LL not seen; isotype, MO).
Similar to O. meziana, but differing in its gla-
brous terminal buds and sharply acute leaf apices.
Collections from Panama have persistent tepals on
the cupule; those from Costa Rica do not.
Distribution. Costa Rica, Panama, from 1300 to
1800 m.
Ocotea whitei Woodson, Ann. Missouri
Bot. Gard. 24: 188. 1937. Nectandra whitei
(Woodson) €. K. Allen, J. Arnold Arbor. 26:
398. 1945. TYPE: Panama. Chiriquí: Seibert
307 (holotype, MO!).
101.
Wrightia 5: 338. 1977. TYPE:
Ocotea eusericea Lundell,
Panama. Chiriquí: Proctor 31858 (holotype, LL not
seen: webs О!
йш? — |. К. ‚ J. Arnold Arbor. 26: 352.
DE — a. Skutch US GH
de seen; isotype, MO
As accepted here, a variable species best rec-
ognized by the dorsally pubescent filaments of the
inner 3 stamens, the rather narrow (2—5 cm wide),
frequently oblanceolate or obovate-elliptic leaves
with a decurrent and sometimes slightly inrolled
base.
Distribution. Nicaragua, Costa Rica, and Pan-
ama, from LOO to 2500 m.
IMPERFECTLY KNOWN SPECIES
102. Ocotea sp. A.
Resembles O. brenesii, from which it differs in
its cymose-paniculate (not racemose) inflorescence,
a denser indument of the the twigs, and a —
(not pubescent) inner surface of the receptacle. 1
is only known from one collection, Carvajal ж
(MO), from Costa Rica. It is likely an undescribed
species, but more material is needed for a descrip-
tion. In Rohwer (1991) it was included as Ocotea
sp. В
P2
Literature Cited
Allen, C. K. 1945. Studies in the Lauraceae VI. A prelim-
inary survey of the Mexican ы Central American spe-
cies. J. Amo — 26: 280-434.
Burger, W. & der Werff.
Costaricensis. qs Idiana, Bof,
Chanderbali, A. S., Н. van der fT & S. Renner. 2001.
Phylogeny and historical — hy of Lauraceae:
Evidence from c — ist and nuclear genomes. Ann.
Missouri Bot. Gard. 88: 104—134.
Gomez-Laurito, J. 199: i new Ocotea (Lauraceae) from
the high ı mountains S Costa Rica and Panama. Novon
313
1990. Lauraceae. In: Fl.
n.s. 23: 1-129.
. 1997. Ocotea morae (Lauraceae): A new species
from Cos a Novon 7: 145-146.
Hammel, В. E. 1986. New species and notes on Lauraceae
from the ( J. Arnold
:aribbean lowlands of Costa Rica.
Arbor. 67: 123-136.
Lorea-Hernandez, F. G. & Н. van der Werff. 2002.
new species of Ocotea * southern Mexico. Brittonia
Three
In press.
Lundell, C. L. 1965. Additions to the Lauraceae of Gua-
temala. — 12: 243—246
————. 1909. Studies of ond Plants I. Wrightia 4:
97-128.
. 1970. Studies of American Plants П. Wrightia 4:
129-1 52.
. 1971. Studies of American Plants Il. Wrightia
4: 153-170.
. 1974a. Studies of American Plants VI. Wrightia
: 23-44.
. 1974b. Studies of American Plants VII. Wrightia
5: 51-7:
— 1977. Studies of American Plants XIV. Wrightia
5: 331-359 А
————. 1978. Studies of American Plants XVI. Wrightia
6: 4—20.
Mez, C. 1889. Lauraceae Americanae. Jahrb. Kónigl. Bot.
art. Berlin 5: 1-550.
Nelson, C. 1984. Una Ocotea (Lauraceae), una Salvia (La-
biatae) y un peara i :ompositae) nuevos de Hon-
duras. Ceiba 25: 173-17€
1986. — einer Monographie der
Gattung Ocotea Aubl. (I — sensu lato. Mitt. Inst.
Allg. puo oe 20: 1-27
Borderline cases 9 ‘tween Ocotea,
<
Nectan-
Volume 89, Number 3 van der Werff 451
2002 Synopsis of Ocotea
dra and Phoebe (Lauraceae): The marginal species of | ———. 1991. A key to the genera of тн in the
the O. helicterifolia group, including the O. heydeana New World. Ann. Missouri Bot, Gard. 78: 377-387.
group. — Jahrb. Syst. 112(3): 365—397. —— ———. |996. Notes on Costa Rican amabas with the
— —. 1993. Lauraceae. Pp. 366—391 in K. Kubitzki, J. dese "s gi se — new species. Novon 6: 476—483
— г & V. Bittrich (editors), The F bem апа. 1999. New taxa and combinations in the Ocotea
E nera of ls ular Plants IL Springer- Verlag. Berlin. helicieiola i auraceae) species group. Novon 9: 571—
van der Wer 1987. Six new — of Neotropical 283. ;
аас Im — Bot. Gard. 74: 401—412 —— 2001. New = and combinations in а
ios eae aie — sal сеае) from Central America. Novon 11: on 511.
[988a. Eight new species sand ue DUM combi- Wendt, T. 1998. Ocotea wa pne (Lauraceae), г pe-
nation of шры Lauraceae. Ann. Missouri Bot.
1
cies of rain forest canopy tree — he md p Te -
Gard. 402— | | — Mexico. Lundellia 1: 40—
1988 88b. A new species of Ocotea (Lauraceae) from є H. van der Werff. 1 = А new spec iés of Ocos
айтарга and a — on n Ocotea jorge-escobarii. Aun. tea (Lauraceae) D Vu Mexico. Ann. Mis-
727. :413—
Missouri Bot. Gard. 7 3-7 souri Bot. Gard.
Volume 89, Number 3, pp. 305—451 of the ANNALS OF m — IRI BOTANICAL GARDEN
is published on October 1, 2002.
—
ANNALS -OF THE MISSOURI BOTANICAL GARDEN
AND MISSOURI BOTANICAL GARDEN ANNUAL REPORT
REE NOW AVAILABLE IN JSTOR!
JSTOR’, a not-for-profit organization, is an important endeavor
dedicated to helping the scholarly community take advantage of advances
in electronic technologies. The JSTOR Archive makes available the
complete backfiles of Annals of The Missouri Botanical Garden,
Missouri Botanical Garden Annual Report (1890-1912), and more than 100
other scholarly journals to researchers through participating libraries.
JSTOR. has recently announced the release of its Ecology & Botany
Collection. Developed in cooperation with the Ecological Society of America,
the Collection includes the archives of twenty-nine journals, offering
online access to more than 1 million pages of important historic literature.
The Ecology & Botany Collection
American Journal of Botany
American Midland Naturalist
American Naturalist
Annals of the Missouri Botanical Garden
Annual Review of Ecology and Systematics
Biotropica
Conservation Biology
Diversity and Distributions
Ecological Applications
Ecological Monographs
Ecology
Evolution
Functional Ecology
Global Ecology and Biogeography Letters
International Journal of Plant Sciences
Journal of Animal Ecology
Journal of Applied Ecology
Journal of Biogeography
Journal of Ecology
Journal of Tropical Ecology
Journal of The Torrey Botanical Society
Limnology and Oceanography
Missouri Botanical Garden Annual Report
New Phytologist
Paleobiology
Quarterly Review of Biology
Systematic Biology
Systematic Botany
Information regarding JSTOR is available at http://www.jstor.org
120 Fifth Avenue, New York, NY 10011
. 3 1753 00294 3147 | 21
|
|
||
!
CONTENTS
Phylogenetic Relationships of the Genus Piptochaetium (Poaceae, Pooideae, Stipeae):
Evidence from Morphological Data
Ana Maria Cialdella & Liliana Ménica Giussani
Systematic Revision and Phylogeny of Paspalum Subgenus Ceresia (Poaceae: Panicoideae:
Paniceae) ............ Silvia 5. Denham, Fernando O. Zuloaga & Osvaldo Morrone
Phylogeny of the Tribe Hymenocallideae (Amaryllidaceae) Based on Morphology and
Molecular Characters
lan W. Meerow, Charles L. Guy, Qin-Bao Li & Jason R. —
Phylogenetic and Biogeographic Diversification in Osmorhiza (Apiaceae) ——
Jun Wen, Porter P. Lowry II, Jeffrey L. Walck & Ki-Oug Yoo
A Synopsis of Ocotea (Lauraceae) in Central America and Southern Mexico —
Henk van der Werff
Cover illustration. | Breonia richardsonii Razafim., drawn by Barbara Alongi.
Annals
of the
Missouri
Botanical
Garden
2002 G
Volume 89, Number 4
Fall 2002
Annals of the
Missouri Botanical Garden
The Annals, published quarterly, contains papers, primarily in systematic botany,
contributed from the Missouri Botanical Garden, St. Louis. Papers originating out-
side the Garden will also be accepted. All manuscripts are peer-reviewed by quali-
fied, independent reviewers. Authors should write the Managing Editor for informa-
tion concerning arrangements for publishing in the ANNALS. Instructions to Authors
are printed in the back of the last issue of each volume and are also available online
at www.mobot.org/mbgpress.
Editorial Committee
Victoria C. Hollowell
ditor,
Missouri Botanical Garden
Amy Scheuler McPherson
Managing Editor,
Missouri Botanical paden
Diana Gunter
Associate Editor,
Missouri Botanical Garden
Aida Kadunic
Senior Secretary
Barbara Mack
Administrative Assistant
Ihsan A. Al-Shehbaz
Missouri Botanical Garden
Gerrit Davidse
Missouri Botanical Garden
Roy E. Gereau
Missouri Botanical Garden
Peter Goldblatt
Missouri Botanical Garden
Gordon McPherson
Missouri Botanical Garden
P. Mick Richardson
Missouri Botanical Garden
Charlotte Taylor
Missouri Botanical Garden
Henk van der Werff
Missouri Botanical Garden
For subscription information contact ANNALS
OF THE MISSOURI GARDEN, % Allen Marketing
& Management, P.O. Box 1897, Lawrence, KS
66044-8897. Subscription price for 2003 is
$145 per volume U.S., $155 Canada & Mexico,
$180 all other countries. Four issues per vol-
ume. The journal Novon is included in the sub-
scription price of the ANNALS.
annals@mobot.org (editorial queries)
http://www.mobot.org/mbgpress
© Missouri Botanical Carden 2002
THE ANNALS OF THE MISSOURI BOTANICAL
GARDEN (ISSN 0026-6493) is published quar-
terly by the Missouri Botanical Garden, 2345
Tower Grove Avenue, St. Louis, MO 63110.
` Periodicals postage paid at St. Louis, MO and
additional mailing offices. POSTMASTER: Send
address changes to ANNALS OF THE MISSOURI
BOTANICAL GARDEN, % Allen Marketing &
Management, P.O. Box 1897, Lawrence, KS
66044-8897
The mission of the Missouri Botanical Garden is to discover and share knowledge about plants and
RD
thei , in order to puan and enrich life.
La k P
© This paper
239.48-1992 (Permanence of Paper).
olume 89 Annals
umber 4: of the
002 Missouri
Botanical
Garden
Ld
PHYLOGENETIC Victor W. Steinmann?
RELATIONSHIPS IN J. Mark Porter?
EUPHORBIEAE
(EUPHORBIACEAE) BASED
ON ITS AND ndhF
SEQUENCE DATA!
and
ABSTRACT
The monophyly and phylogenetic relationships of the tribe Euphorbieae (Euphorbiaceae) were evaluated using sep-
juences of the nrDNA internal transc ribed spacer (ITS)
sec
s representing nearly all of the previously
arate — maximum parsimony analyses of nucleotide
region and cpDNA coding region ndAF. The study included 223 ingroup speci
recognized genera, subgenera, and sections within the К uphorbieae and 4 outgroup taxa from the tribe Hippomanose
(E прогини eae е). Te th the ITS and ndhF analyses support the monophyly of Euphorbieae in addition to the monophy
subtribes. Anthosteminae, Neoguillauminiinae, and Euphorbiinae. Within Euphorbiinae, there are four major
vhich copos ds w with a previously recognized taxon: the three remaining clades
d of a paraphyletic ety All other
ind Endadenium
of its three
clades, — one of y
of various subge nera and sections. The majority of the subtribe is compose
genera currently recognized in the subtribe are nested within Euphorbia. In addition, Synadenium ai
are nested within — nium. Within Euphorbia, the majority of the currently recognized subgenera are either par-
aphyletic or polyphyle vlet iogeographical patterns examined in light of the also idence suggest that the tribe
Euphorbieae arose in Africa. possibly before the breakup of Gondwanaland, at which time aha major — s of subtribe
^A
| are extremely grateful to all those who have assisted with this research, In partic ‘ular we thank Susan Carter,
J. Travis Columbus, Elisabeth Friar, Victoria Hollowell, Gordon McPherson, and Grady Webster for critically reviewing
the — ap For — to e — their collections and to sample he еа material, we thank the curators
. DAV. ENCB. E GB. K. MO. NY. RSA. SGO. SP, and TEX. The following institutions and
and s ARIZ, C.
iid extended assistance we pert o access a their living collections: Dylan Hannon; Chuck Hanson of Arid
Lands Greer — ‘sı Mark Mayfield; Meena Singh: mour Linden; John Trager and Joe Clements of the Huntington
Botanic Garden: Urs Egeli o of с, anten- — Zürich: and Herman Schwartz. Caden McPherson and Sebastian
Teillier collecte d material of gemini cleopatra and Euphorbia germainii, respectively. Fieldwork, herbarium
visits, and laboratory use and mate ‘rials were made possible by the generous support of the Andrew W. Mellon Foun-
dation, the American Society of Plant | жикке, m Howard and Phoebe Brown Sc holarship, and the е
Native Plant Research Award. Lauren Raz provided ITS sequences for C иаа degeneri, C. hypericifolia, a
rata. Kenneth Wurdack gave useful information about outgroup selection. We are inde bted to the staff and ——
Garden for encouragement and support, especially Eric Roalson for his
=
prost
students at the Rancho Santa Ana Botanic
Е help апа advi
inc үш; — Ana Botanic Garden,
" 1.Mx
1500 N. College Ave.. Claremont, California 91711, U.S.A. steinmann@
‘Tine — Instituto de Ecología. A.C., Centro Regional del Bajío, A.P. 386, 61600 Pátzcuaro, Michoacán,
Mexico.
ANN. Missouri Bor. GARD. 89: 453—490. 2002.
454
Annals of the
Missouri Botanical Garden
Euphorbiinae were alre
scription of Euphorbia that contains all of
Key words:
the about 20(
sady present. It is argued that the best solution for Euphorbia classification is a broad circum-
JO species of the subtribe Euphorbiinae.
classification, Euphorbia, Euphorbieae, ITS, ndhF.
It was only a year after Linnaeus (1753) first cir-
cumscribed the genus Euphorbia L. (Euphorbi-
aceae) to include all of the then-known members
of the tribe Euphorbieae that other botanists began
to divide the genus into several smaller genera
(Miller, 1754;
continued since, and almost 250 years later the is-
sue as to whether Euphorbia should be recognized
Trew, 1754). The controversy has
in its initial broad sense or be separated into many
smaller genera still has not been resolved. Euphor-
bia and Euphorbieae are generally considered tax-
onomically difficult, and a considerable degree of
uncertainty has always existed about the relation-
ships of the groups within them. This problem is
due in great part to extreme morphological diver-
sity, a large number of species, and a subcosmo-
politan distribution. Few workers have been able to
gain a complete understanding of the tribe through-
out ils immense range, and there has never been a
This lack of
consistency has hindered and to some extent dis-
couraged research within the genus Euphorbia as
well as the
universally accepted classification.
Euphorbieae as a whole. Thus, a clear
understanding of relationships within the tribe is of
great importance in order to provide subsequent
workers with a phylogenetic framework on which to
base their studies.
The tribe Euphorbieae is characterized by its
synapomorphic pseudanthial inflorescence (termed
a cyathium) composed of a gland-bearing involucre
of several united bracts and their associated flowers
and bracteoles. Each bract subtends a monochasial
staminate inflorescence, and these monochasia sur-
round a single pistillate flower. The individual flow-
ers in Euphorbieae are highly reduced and repre-
sented by a single stamen or ovary, with or without
a perianth. The staminate flowers terminate slender
pedicels, and the pistillate flowers can be long-ped-
icellate or subsessile. This structure is highly com-
plex, and there are still doubts as to its exact nature
and evolution (Gilbert, 1994).
According to the most recent synopsis of the Eu-
phorbiaceae (Webster, 1994), the Euphorbieae con-
tain 11 genera placed into three subtribes: Anthos-
teminae (Baill.) G. L. Webster,
Croizat, and Euphorbiinae.
Neoguillauminiinae
r ab-
sence of a perianth (presumably a calyx) on the
flowers distinguishes these taxa. In Anthosteminae
both the staminate and. pistillate flowers possess a
The presence «
perianth; in Neoguillauminiinae only the pistillate
flowers possess a perianth; and in Euphorbiinae
neither the pistillate nor the staminate flowers pos-
sess a well-developed perianth, although a rudi-
mentary calyx-like structure is present below the
pistillate flowers in a few species.
Anthosteminae consist of two genera of tropical
forest trees: Anthostema А. Juss. (4 spp. disjunct in
west tropical Africa and Madagascar) and Dichos-
temma Pierre (1 sp. in west tropical Africa). This
subtribe is considered the least specialized because
of the perianth on both the staminate and pistillate
Further, each individual involucral bract
closely envelops a cluster of many bracteoles and
staminate flowers.
flowers.
The cyathia are bisexual or sta-
minate. In Anthostema the cyathia are arranged in
condensed axillary cymes. The involucre is made
up of four united bracts and is slightly zygomorphic
because it is spread in an open half circle. There
are five large glands along the involucral bract mar-
gins and between the clusters of staminate flowers.
The 3-locular pistillate flower is not contained in
the involucre but instead lies at the base of the
involucre’s open side. Based on its open involucral
morphology and the possible lateral position of the
pistillate flower, this genus probably most closely
resembles the ancestral inflorescence morphology
of the Euphorbieae. In Dichostemma the cyathia are
arranged in loose, open, axillary or terminal cymes.
In contrast to the four monochasia-
containing involucral bracts are united in a
Anthostema,
ring,
and the involucre is completely closed and acti-
nomorphic. The four involucral glands are con-
tained. within this structure and attached to the in-
ner walls of the and the base of the
gynophore. In the center of the inflorescence is ei-
ther a minute pistillode or a 4-locular pistillate
flower.
bracts
Like Anthosteminae, the subtribe Neoguillau-
miniinae also contains two genera: Calycopeplus (5
spp. in Australia) and Neoguillauminia Croizat (1
sp. in New Caledonia). In contrast to Anthostemi-
nae, the staminate flowers lack a perianth. How-
ever, a perianth is present on the pistillate flower,
and this combination of features defines the sub-
tribe. In addition, the bracts of the involucre do not
tightly envelop the staminate monochasia, although
the latter are enclosed within large bracteoles. Ca-
lycopeplus are xerophytic shrubs with small, oppo-
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
site leaves and cyathia arranged in axillary clusters.
The involucre is distinctly cupular and is made up
of four bracts. The glands are located between the
lobes and attached to the wall of the cupular in-
voluere. There is a central pistillate flower sur-
rounded by four staminate monochasia that are op-
posite the involucral lobes. The sole representative
of Neoguillauminia is a mesic forest tree or shrub
with large, spirally arranged leaves. The cyathia are
long-pedunculate and arranged in few-cyathiate ax-
illary or subterminal groups. The involucre is com-
posed of four to six bracts that extend into large,
petaloid appendages. There are eight to twelve
glands arranged in pairs between the four to six
staminate monochasia and attached to the base of
the involucre and sometimes also to the base of the
gynophore. There is a single central pistillate flow-
er. Neoguillauminia is noteworthy because the in-
volucres are generally composed of five bracts, and
a 5-merous involucre also characterizes Euphorbi-
inae, discussed below. Calycopeplus, as well as An-
thostema and Dichostemma, possess 4-merous in-
volucres
With about 2000 species and a subcosmopolitan
distribution, the largest and most complex subtribe
is Euphorbiinae. It is characterized by the lack of
a perianth on both the staminate and pistillate flow-
ers, although a rudimentary calyx-like structure is
present below the pistillate flowers of a few species.
In addition, the involucre is made up of five united
bracts, not four as generally are found in the other
subtribes of Euphorbieae, and the bracteoles that
surround the staminate monochasia are generally
reduced. The glands are mostly located along the
rim of a cupular involucre. Again following the cir-
cumscription of Webster (1994), the subtribe con-
tains seven genera: Chamaesyce Gray, Cubanthus
(Boiss.) Millsp., Endadenium Leach, Euphorbia.
Monadenium Pax, Pedilanthus Necker ex Poit., and
Synadenium Boiss. Poinsettia Graham and Elaeo-
phorbia Stapf are also sometimes recognized as dis-
tinct from Euphorbia, but their status is not as
widely accepted, and most authors treat these as
infrageneric taxa of Euphorbia. Extreme emphasis
is placed on variation in the configuration of the
eyathium, and with the exception of Chamaesyce,
the genera segregated from Euphorbia are distin-
guished on the basis of involucral features.
Euphorbiinae are dominated by Euphorbia,
which accounts for ca. 80% of the species and oc-
curs throughout the geographic range of the sub-
tribe. The genus is best known for the common
Christmas poinsettia (Euphorbia pulcherrima Willd.
ex Klotzsch) and is popular with horticulturists be-
cause of the prevalence of succulents. One of the
most fascinating features of the subtribe is its great
diversity of growth forms. Large forest trees, shrubs,
perennial herbs, geophytes, annuals, and a great
diversity of succulents are all well represented. In
comparison to most other genera of Euphorbiinae.
the cyathial morphology of Euphorbia is relatively
unspecialized. The cyathia are actinomorphic and
generally possess one to five separate glands situ-
ated on the rim of the involucre, and this plesiom-
orphic feature unites the genus.
With about 300 species, Chamaesyce is the larg-
esl segregate genus from Euphorbia. lt occurs wide-
ly, but most species are confined to the New World.
Distinguished on the basis of vegetative morphol-
ogy, its cyathia are nearly identical to those of many
species of Euphorbia subg. Agaloma (Raf.) House.
Chamaesyce is characterized by many unusual syn-
apomorphies: apical abortion of the main shoot and
subsequent. sympodial growth; interpetiolar stip-
ules: opposite, frequently asymmetrical leaves; and
C, photosynthesis (Koutnik, 1984. 1987). Numer-
ous Euphorbieae specialists (e.g.. Carter, 1988a,
1992b: 1987) retain it as a subgenus of
Euphorbia.
Gilbert,
The other five segregate genera of Euphorbia are
also easily identifiable, but their differences, as
mentioned above, involve involucral features. Three
genera possess zygomorphic cyathia: in Pedilanthus
primarily Mexico) there are two to six
glands enclosed within an adaxial, spurlike exten-
sion of the involucre; in Cubanthus (3 spp., Cuba
and Hispaniola) there are two glands united into a
(15 spp..
shieldlike structure on the outside of the involucre:
and in Monadenium (ca. 70 spp., Africa) the glands
are united into a single horseshoe-shaped structure.
The two remaining segregate genera possess acti-
nomorphic eyathia: in Synadenium (20 spp., Africa)
the eyathia possess five united glands that form a
complete ring around the top, and in Endadenium
(1 sp.. Angola) the closed rim of the cyathium is
not a gland but instead an apparently eglandular
extension of the involucral wall with a ring of nec-
tar-bearing depressions on the inside of the invo-
lucre
TAXONOMIC HISTORY OF THE TRIBE EUPHORBIEAE
The taxonomic works treating Euphorbieae are
numerous, and only a brief overview, focusing on
those with broad and significant implications to
modern Euphorbieae taxonomy, will be provided
here. A more detailed account is found in Stein-
mann (2001
One of the first prominent taxonomic treatments
—
of the Euphorbieae was that of Klotzsch and Garcke
456 Annals of the
Missouri Botanical Garden
(1859, 1860). Although these authors segregated that of Pax and Hoffmann (1931) in Engler’s Die
both Anthostema and Pedilanthus into separate
tribes, the Anthostemeae and Pedilantheae, they
did concede their close relationship with Euphor-
bieae. Their Euphorbieae contained 408 species,
all members of present-day Euphorbiinae. Euphor-
bia was recognized in a restricted sense, with only
27 species. The remainder were placed in 17 seg-
regate genera, organized in two subtribes: Aniso-
phyllae Klotzsch & Garcke
Klotzsch & Garcke.
The first and last monograph of the entire Eu-
and Tithymalae
phorbieae, and the basic framework for the classi-
fication that is still in use today, was provided by
(supplement 1866) for de Can-
Boissier followed Klotzsch and
Boissier in 1862
dolle's Prodromus.
Garcke by recognizing Anthostemeae as distinct,
but differed from them by submerging Pedilantheae
into Euphorbieae. Euphorbieae contained 740 spe-
cies. Also in contrast to Klotzsch and Garcke, Bois-
sier recognized Euphorbia in a broad sense, and
his Euphorbieae contained only three genera, all
currently placed in Euphorbiinae: Pedilanthus (15
spp.). Synadenium (2 spp.), and Euphorbia (723
spp.). The taxa of Euphorbia were positioned into
27 sections and two “series”: Appendiculatae
(corresponding to Klotzsch and Garcke's
Anisophyllae) and
(corresponding to Klotzsch and Garcke's
Boiss.
subtribe Exappendiculatae
Joiss.
subtribe Tithymalae). Calycopeplus and Neoguillau-
minia were treated in Euphorbia, and Cubanthus
was erected as a section of Pedilanthus.
Bentham (1878) expounded greatly on the Eu-
phorbiaceae as a prelude to his treatment of the
family in Genera Plantarum (Bentham, 1880). He
agreed that Euphorbia should be recognized in the
broad sense and followed Boissier’s treatment with
only minor modifications. Anthostemae was com-
bined with Euphorbieae. Calycopeplus, first de-
seribed by Planchon in 1861 but subsequently
treated as a synonym of Euphorbia by Boissier
(1862), was resurrected at the rank of genus. Ben-
tham’s greatest change concerned the rank of Bois-
sier's sections of Euphorbia, and he proposed a sys-
tem containing only six sections, under which the
majority of Boissier’s sections were reduced to sub-
sections. Otherwise, Bentham did little to modify
Boissier's grouping of species.
Around the turn of the century, Pax (1894a) de-
scribed Monadenium and Pierre (1896) described
Dichostemma, the second known genus in the mod-
Nearly two decades
later, Millspaugh (1913) first recognized Cubanthus
ern subtribe Anthosteminae.
as a distinct genus.
The next major treatment of the Euphorbieae was
natürlichen Pflanzenfamilien. They essentially fol-
Bentham's classification with only minor
newly
taxa. Again, a single tribe, the Euphorbieae, was
Owe
o accommodate described
e
modifications
recognized without any further divisions. Anthoste-
ma, Dichostemma, and Calycopeplus were all treat-
ed as distinct. Euphorbia continued to be recog-
nized in the broad sense, although the number of
segregate genera had increased. In addition to the
earlier recognized Pedilanthus and Synadenium,
these authors segregated the genera Monadenium,
Stenadenium Pax, Elaeophorbia, and Diplocy-
athium H. Schmidt.
now treated within Euphorbia (Webster, 1994), and
Stenadenium is currently — as a synonym of
Monadenium (Bally, 195°
relegated back to a section of Pedilanthus.
The latter two segregates are
. 1961). Cubanthus was
In 1937, Croizat described the genus Neoguil-
lauminia and the subtribe Neoguillauminiinae to
accommodate it. The type, N. cleopatra, was first
described by Baillon (1861) as a species of Eu-
phorbia and placed in his monotypic E. sect. De-
cadenia Baill. Boissier (1862) treated this species
within E. sect. Tithymalus.
Wheeler
(1943) published a broad classification of the entire
Shortly after. Croizat's contribution,
Euphorbieae. This was the first major conspectus
to advocate the use of subgenus as the primary di-
vision of Euphorbia, and it is this rank that pre-
dominates in current. Euphorbiinae classification.
Wheeler recognized a single Euphorbieae but
abandoned Croizat's subtribes. The genera that he
included were the same as those of Pax and Hoff-
mann (1931) except that Diplocyathium was rele-
gated to synonymy within Euphorbia, and Croizat's
newly described Neoguillauminia was recognized.
Euphorbia consisted of eight subgenera, and in
general, these corresponded to the sections recog-
(1931).
Dressler (1957) provided a monograph of Pedi-
nized by Pax and Hoffmann
lanthus in which he convincingly demonstrated that
the species of Cubanthus do not belong within Ped-
ilanthus and should be treated separately. Dressler
is also noteworthy because he advocated that Eu-
phorbia should be recognized in the narrow sense
and restricted to Old World succulents.
In the years after Dressler's treatment, Webster
made important contributions to Euphorbieae tax-
onomy. He suggested the recognition of Chamae-
syce as a genus and at the same time strongly sug-
gested that Elaeophorbia should be treated within
Euphorbia (Webster, 1967). Eight years later, Web-
er (1975) circumscribed the tribe to its present,
generally accepted. configuration. He created. the
Volume 89, Number 4
2002
Steinmann & Porter 457
Phylogenetic Relationships in Euphorbieae
subtribe Anthosteminae to accommodate Anthoste-
ma and Dichostemma and resurrected Neoguillau-
miniinae to include Calycopeplus and Neoguillau-
minia. In his third subtribe Euphorbiinae, he
recognized seven genera: Euphorbia, Chamaesyce,
Cubanthus, Endadenium, Monadenium, Pedilan-
thus, and Synadentum.
Gilbert (1987) and Carter (1985, 1988a) have
made the most recent modifications in Euphorbi-
inae classification. Their changes primarily involve
African members and the elevation of various pre-
viously recognized sections of Euphorbia to sub-
generic rank
In summary, at the time of this writing most au-
thors agree that Euphorbieae is a single tribe with
three subtribes: the Anthosteminae, the Neoguillau-
miniinae, and the Euphorbiinae. Within Anthos-
teminae there are two genera, Anthostema and Di-
chostemma. Within Neoguillauminiinae there are
also two genera, Calycopeplus and Neoguillaumi-
nia. The subtribe Euphorbiinae contains about sev-
en genera. Those that are universally accepted are
Endadenium, Monadenium,
Euphorbia, Synaden-
ium, Pedilanthus, and Cubanthus. Genera that are
less frequently accepted are Chamaescye, Elaeo-
Other
have not even gained minor acceptance. Within Fu-
phorbia, and Poinsettia. segregale genera
phorbia, the common primary division employed is
the rank of subgenus, and there are 9 to 11 gen-
erally recognized subgenera: Agaloma, Chamaesyce
Raf. (when not treated as a genus), Poinsettia (Gra-
ham) House (when not treated as a genus), Esula
Pers., Eremophyton (Boiss.) 7 C. Wheeler, Fu-
phorbia, Lacanthis (Raf.) M. G. Gilbert, Tirucalli
(Boiss.) S. Carter, Trichadenia pee S. Carter, Rhi-
zanthium (Boiss.) L. C. and Lyctopsis
(Boiss.) L. С.
coherent classification of the genus is lacking. and
heeler,
Wheeler. However, a modern, global,
some proposed sections of Euphorbia have not been
adequately accommodated within a currently rec-
ognized subgenus (e.g., E. sects. Arthrothamnus
(Klotzsch & Garcke) Boiss.,
dri) Croizat, and Deuterocalli Croizat). Detailed ac-
Denisophorbia (Lean-
counts of the currently recognized infrageneric taxa
of Euphorbia and their taxonomic histories are pre-
sented in the discussion section.
MATERIALS AND METHODS
In total, 227 species were sampled (Appendix 1),
including 4 outgroup and 223 ingroup species. Out-
groups were chosen from tribe Hippomaneae be-
cause it is traditionally considered to be closely
related to the Euphorbieae (Webster, 1994
broad molecular phylogenetic reconstruction of the
. and a
м
entire Euphorbiaceae also supports their affinity
(Kenneth Wurdack, pers. comm.). An attempt was
made to include as broad a sample as possible from
Euphorbieae. Four species of Anthosteminae, 4
species of Neoguillauminiinae, and 215 species of
Euphorbiinae were sequenced. These included all
genera (except Cubanthus) recognized by Webster
1994).
eight species of Pedilanthus, one species of Syna-
The sole representative of Endadenium,
—_
denium, eight species of Chamaesyce, and five spe-
cies of Monadenium were included. One hundred
ninety-two species of Euphorbia were represented,
including all sections (except sects. Bongium Boiss.
and Caulanthium Boiss.) treated by Boissier (1862)
and most other important taxonomic groups recog-
nized by subsequent workers (e.g., Webster, 1967;
Gilbert, 1987; Carter, 1985, 1988a).
The ITS
3 Anthosteminae, 2 Neoguillauminiinae,
analysis included 216 species (4 out-
groups.
and 207 Euphorbiinae). The sample for the ndhF
analysis was smaller and included 114 species (4
oulgroups, 4 Anthosteminae, 3 Neoguillauminiinae,
and 103 Euphorbiinae). Following a preliminary
analysis of the ITS sequence data, a subset of the
sampled taxa representing the major clades and
well-supported lineages was sequenced for the
ndhV analysis. In addition, 15 species were includ-
ed in the ndhF analysis that were not included in
the ITS analysis because of problems obtaining
“clean” ITS sequences for these taxa (see Appen-
dix 1).
Total DNA was isolated from either
fresh, silica gel-dried, or herbarium material using
a modified CTAB method (Doyle & Doyle, 1987).
Two genic regions were employed in the phyloge-
genomic
netic. reconstructions: the cpDNA coding region
ndh V
region (ITS). ITS amplification using the polymer-
and the nrDNA internal transcribed spacer
ase chain reaction (PCR) followed the procedures
described by Baldwin (1992) and Baldwin et al.
1995). Amplification of the ndAF region generally
—
followed the protocols described by Olmstead and
Sweere (1994) and Kim and Jansen (1995). The 5’
quarter of the ndhF region was excluded due to
problems amplifying it. Also, a primer ca. 50 bp
internal to the 3^ end “2110Ri” (5'-TCA ATT ATT
CGT TTA TCA А-3”) was designed because many
taxa would not amplify using primer “2110R.” Four
additional primers were specifically designed for
this study: (1) (5'-FTA TTC AAT АТС TYT ATG
GGG TAA-3’), (2) (5'-TAA CCC CAT АҢА GAT
ATT GAA TAA-35, (3) (5'-TAG GAA TTC CYT
TYA ATC AA-3'), and (4) (5'-TTG ATT RAA RGG
AAT TCC TA-3’).
The PCR products were electrophoresed using a
458
Annals of the
Missouri Botanical Garden
Tabl
Table 1
Characteristics of the ITS and portion of the ndhF regions included in this study.
ITS ITS] 5.85 ITS2 ndhF
Raw length (bp) 591-660 210-267 164-167 202-241 1467-1506
Aligned length 739 303 169 267 587
Variable sites (proportion) 527 (0.71) 272 (0.90) 35 (0.21) 220 (0.82) 740 (0.47)
Parsimony-informative sites
=
70 (0.64) 244
(proportior
GC content, mean (range)
( ((
0.58 (0.50-0.70) 0.58 (0.46-0.71) 0.56 (0.51—0.58) 0.61 (0.48-0.76)
22 (0.13) 204 (0.76) 519 (0.33)
0.32 (0.30-0.33)
1.5% agrose gel in a 0.5x TBE (pH 8.3) buffer,
stained with ethidium bromide, and then cleaned
using the PEG precipitation. protocol (Nickrent,
1996). Cycle-sequencing adhered to the manufac-
turers specification using the PRISM®
Deoxy™ Terminator Kit (Perkin-Elmer,
cle-sequencing
)ye-
Inc.). Cy-
was followed by ethanol
purification, and sequencing used an Applied Bio-
systems Model 373A Automated DNA Sequencing
System.
Sequences were assembled from automated DNA
3.0
TS sequences
sequence chromatograms using Sequencher
Inc.).
were initially aligned with ClustalW v. 1.4 (Thomp-
(Gene Codes Corporation,
son et al., 1994), using a gap costgap extension
cost ratio of 10:5, followed by visual modifications;
ndhF sequences were aligned visually. Because of
high divergence and the large number of taxa in-
cluded in the study, alignment was problematic for
certain highly variable regions of ITS sequences.
However, the difficulties mostly occurred aligning
the major lineages of Euphorbieae to each other,
and alignment within major lineages was less prob-
and ndhF
duced gaps into some sequences, and these sites
lematic. Alignment of both ITS intro-
Missing data were
coded with a question mark in the matrix. Align-
were included in the analyses.
ment matrices have been deposited at the library
of Rancho Santa Ana Botanic Garden and submit-
o TreeBASE
treebase.index.html).
e
ted (http://herbaria.harvard.edu/
The aligned ITS and ndhF sequence matrices
were analyzed separately using PAUP* 4.0b4a for
Macintosh™ (Swofford, 2000) on a Macintosh™ G3.
Due to the large data sets, maximum parsimony
using heuristic searches (Acctran, 10 random ad-
dition cycles, TBR branch swapping, steepest de-
scent option not in effect) was employed. Maximum
likelihood estimates of transition/transversion (TI/
TV) biases were measured for both the ITS and
ndhk data sets individually as implemented in
PAUP* 4.0b4a under the HKY model of nucleotide
substitution and using the equal-weighted parsi-
mony trees. These estimates were used to produce
a TI/TV step matrix, employed in further maximum
parsimony analyses. This weighting scheme was
employed in order to model more closely the max-
imum parsimony analyses to the given data set. In-
dels were coded as missing data. Multiple most par-
trees combined in a strict
simonious were
consensus tree. Tree robustness was estimated us-
ing 10,000 “fast addition” heuristic bootstrap rep-
licates.
RESULTS
Sequence variation for both ITS and ndhF
shown in Table 1. For ITS sequences, pairwise lev-
” values) for the
34.4% (between
Anthostema sp. nov. and Euphorbia insulana) to
1.1% (between E. alta and E. spathulata). For sub-
els of divergence (uncorrected *p
entire Euphorbieae ranged from
tribe Anthosteminae, levels of divergence varied
from 9.4% (between Dichostemma glaucescens and
Anthostema madagascariense) to 2.2% (between A.
madagascariense and A. sp. nov.); for the two in-
cluded species of subtribe Neoguillauminiinae
(Neoguillauminia cleopatra and Calycopeplus cas-
uarinoides), the level of divergence was 7.696; for
subtribe Euphorbiinae levels of divergence varied
from 30.9% (between E. trichotoma and E. panch-
ganiensis) to 1.1% (between E. alta and E. spathu-
ata). As expected, levels of divergence for ndhF
sequences were much lower than levels observed
in. ITS
these ranged from 11.796 (between Anthostema sp.
sequences. For the entire Euphorbieae,
nov. and E. oaxacana) to 0.1% (between E. bilobata
and E.
levels of divergence varied from 2.696 (between An-
exstipulata). For subtribe Anthosteminae,
thostema sp. nov. and Dichostemma glaucescens) to
0.496 (between Anthostema sp. nov. and A. mada-
gascariense); for subtribe Neoguillauminiinae di-
vergence levels varied from 296 (between Calyco-
peplus collinus and C. paucifolius) to 1.496 (between
Neoguillauminia cleopatra and Calycopeplus pau-
cifolius); for subtribe Euphorbiinae levels of diver-
gence varied from 9.6% (between E. aphylla and
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
459
Table 2
weighted characters of the same topology. For e
Statistics from weighted maximum ee analyses of ITS and ndhF datasets and statistics for un-
analysis, number and length of most parsimonious trees, consistency
index (CI). retention index (RI), and rescaled consistency index (RC) are reported.
Number
Analysis of trees Tree length Cl RI RC
ITS (weighted) 2160 8838.2 0.1813 0.7219 0.1309
ITS (unweighted) 2160 6887 0.1802 0.7126 0.1284
ndhF (weighted) 19012 1818.2 0.5221 0.7926 0.4138
ndhF (unweighted) 19012 1985 0.5441 0.7992 0.4349
E. oaxacana) to 0.1% (between E. bilobata and E.
exstipulata).
Transition/transversion biases were calculated to
be 1.8 for the ITS data set and 0.85 for the ndhF
data set. These biases were used in the weighted
maximum parsimony analyses.
The ITS analysis resulted in the recovery of 2160
trees of 6887 steps (equal
most parsimonious
weighted). The ndhF analysis resulted in the re-
covery of 19,012 most parsimonious trees of 1985
steps (equal weighted). Statistics for both analyses
are detailed in Table 2. The strict consensus of the
2160 trees obtained in the ITS analysis is depicted
in Figures 1, 5. 7, 9, 10, 12, and 13. The strict
consensus of the 19012 trees obtained in the ndhF
analysis is depicted in Figures 2, 6, 8, 11. and 14.
With the exception of Figures 3 and 4. phylograms
are presented in Steinmann (2001).
For the ITS analysis, the bootstrap 50% majority
rule tree (values depicted on strict consensus tree)
is structurally identical to the strict consensus in-
sofar as the majority rule tree is resolved. Many of
the terminal clades are supported with high boot-
strap (bs) percentage values. However, there is no
support for the majority of the basal internal nodes.
Notable exceptions are the ancestral nodes that de-
fine the tribe Euphorbieae and its subtribes. In the
ndhV analysis, the bootstrap 50% majority rule tree
(values depicted on strict consensus tree in Fig. 2)
is structurally identical to the strict consensus with
one significant difference. In the bootstrap analysis,
clade A is not placed as the sister taxon of clade
B but instead is located at the earliest diverging
clade in a group that contains clades C and D. and
this is supported by a bootstrap value of 77%. This
incongruency suggests that the topology of the ma-
jor clades within the subtribe Euphorbiinae should
be viewed with caution. The majority of clades
found in the ndhF strict consensus tree are sup-
ported in the bootstrap analysis, and in contrast to
ITS (see Fig. 1). many of the internal, basal nodes
also possess significant bootstrap support.
It is worth noting that a combined analysis was
conducted for the 99 species in common between
the ITS and ndhF data sets.
presented here because results from the indepen-
dent ITS and ndhF analyses agree strongly with
each other, and the combined analysis does not pro-
vide novel insights. For the instances in which tree
structure resulting from the ITS and ndhF analyses
differed, the combined analysis generally provided
The results are not
the same structure as the ndhF analysis.
DISCUSSION
UTILITY OF THE ndhF AND ITS REGIONS IN
EUPHORBIACEAE AND A COMPARISON OF THE
RESULTING PHYLOGENIES
Despite being one of the largest angiosperm fam-
ilies and an important floristic Component. espe-
cially in tropical regions, the Euphorbiaceae have
not previously been the subject of a comprehensive
molecular systematics investigation. If the results
obtained here are any indication, both ndhF and
the ITS data appear useful in inferring relationships
within the family. In general, the two reconstruc-
tions of the Euphorbieae are similar, and thus cor-
roborate evidence of their phylogenetic signal. The
— structure of both analyses is the same (Figs.
1, 2). and when differences do occur, these gener-
ally involve clades that lack bootstrap support in
one or both of the There are very few
instances in which well-supported clades are po-
sitioned differently in the separate analyses, e.g.,
the arrangement of Euphorbia meenae Blatt. &
McCann, E. abdelkuri Balf.f., E. drupifera Thonn..
and E. poissonii Pax (Figs. 10, 11)
analyses.
MONOPHYLY OF THE EUPHORBIEAE AND ITS
SUBTRIBES AND PARAPHYLY WITHIN SUBTRIBE
EUPHORBIINAE
In both analyses, Euphorbieae form a monophy-
letic assemblage with respect to the four outgroups.
However, there is not high bootstrap support for this
grouping in either analysis. The unusual synapo-
460
Annals of the
Missouri Botanical Garden
Stillingia spinulosa
Figure 1.
Bootstrap — greater than 50% are shown above the
Webster (1994) are indicated. Individual strict consensuses for c Jade 's A-D are found in Figures 5, 7, 9, 10, 12, ar
morphic inflorescence structure of the tribe there-
ore supports this relationship, and it is highly
probable that the tribe is indeed monophyletic. All
three subtribes of Euphorbieae also form monophy-
letic groups (Figs. 1, 2), and these clades possess
high bootstrap support, 100% for all three in the
ndh¥ analysis. Further, in both analyses subtribe
Anthosteminae is the earliest diverging subtribe in
Euphorbieae and subtribes Neoguillauminiinae and
Euphorbiinae are sister to each other. Anthostema
and Calycopeplus are strongly supported to
monophyletic (bs 100% and 90% in ndhF, respec-
tively), and presumably so too are the monotypic
Dichostemma and
—
w
^
Neoguillauminia. Therefore,
clade D
(Widespread)
=
clade C ie
82 (Widespread) 30
ЕЁ
clade В zm
(Primarily Laurasia) "
95 92
— clade A tr
(Africa and Madagascar) С
‚ 0
, 5 T
Calycopeplus casuarinoides ai 210
100 (Australia) 220 55
La | W
67_ Neoguillauminia cleopatra 7 >т A
(New Caledonia) z т
d madagascariense | >
100 (Madagascar) —
Ic
gc
Anthostema sp. nov. Ax
100 mA
(Madagascar) ez
Z 1]
>
Dichostemma glaucescens т
(Tropical West Africa)
Sapium sebiferum
100
— ©
— — Omalanthus populifolius С
—
(2
*
Е ©
Sebastiania cornuta ©
98 GF
#2]
Strict consensus of 2160 trees based on a weighted maximum pamimony analysis of the ITS region.
; of the classification of
nd 13.
branches. The tribes and subtribes
Webster's overall classification of the tribe Euphor-
bieae (1975) is consistent with the molecular re-
sults.
Within Euphorbiinae, both analyses demonstrate
that Chamaesyce, Pedilanthus, Monadenium, Syn-
adenium, and Endadenium are all nested within a
paraphyletic Euphorbia (see clades C and D, Figs.
9-14). Also, both Synadenium and Endadenium are
nested within a paraphyletic Monadenium.
RELATIONSHIPS AND MAJOR LINEAGES WITHIN THE
EUPHORBIINAE
Both analyses support that there are four major
lineages within the subtribe Euphorbiinae. here
Volume 89, Number 4 Steinmann & Porter
2002 Phylogenetic Relationships in Euphorbieae
74
clade D
91 (Widespread)
m
75. elade С 32
100 (Widespread) © E
86 F
clade B ЫШ
(Primarily Laurasia) >
100
100 clade A
— (Africa and Madagascar)
2
Calycopeplus collinus © т
90 (Australia) а s c
Е S az
100 Co alye е paucifolius S =
(Australia en} ©
zz | x
62 Neoguillauminia cleopatra = =
(New Caledonia) > >
"Е
Anthostema madagascariense
78 (Madagascar) >
2,
du
100 Anthostema 2 nov. Е E
(Madagascar ny
=E
100 Anthostema senegalense Sm
(Tropical West Africa) >
Dichostemma — escens
(Tropical West Afri
Sapium sebiferum
Omalanthus populifolius
Sebastiania cornuta
SdNOaADLNO
Stillingia spinulosa
Figure 2. Strict consensus of 19,012 trees based on a weighted maximum рагайдопу analysis of the ndhF region.
Bootstrap values greater than 50% are shown above the branches. The tribes and subtribes of the — ‘ation of
Webster (1994) are indicated. Individual strict consensuses for dades A-D are found in Figures 6, 8, 11, and 14.
Due to the lack of a comprehensive classifica-
designated clades A-D. With the exception of clade
tion, many sections of Euphorbia have not been ac-
B (primarily composed of the temperate herbaceous
group of Euphorbia subg. Esula), none of these lin- commodated within a currently recognized subge-
eages closely correspond to any previously recog- nus. Also, some sections that have been suggested
nized taxon. Instead, they are conglomerates of var- to belong to a particular subgenus are demonstrated
ious sections and subgenera. All of these major to not be closely related. Therefore, the following
lineages possess significant bootstrap support in the discussion about the major groups of Euphorbiinae
ndhV analysis (bs 74-100%), while only clade А is is organized by taxa of various ranks.
supported in the ITS analysis (bs 92%). There still
is some question, however, as to the exact relation-
ship among these lineages. Although both the ITS were included. In both the ITS and ndhF analyses
and ndhF strict consensus trees show that clades (Figs. 12, 14), these species form a well-supported,
A and B are sister to each other and together these monophyletic group (bs 91% and 99%, respective-
are sister to clades C and D, there is no bootstrap ly) in clade D. Species previously thought to inter-
support to this grouping in either analysis. grade with Chamaesyce (see Webster, 1967), such
Chamaesyce. Eight species from Chamaesyce
462 Annals of the
Missouri Botanical Garden
Е. gradyi
E. guerichiana | clade D
E. й
Е. tanguahuete саде С
E. trichotoma
E. calyptrata | clade B
E. acalyphoides
] clade A
copeplus casuarinoides
eoguillauminia cleopatra
lense
v.
Dichostemma glaucescens
Sapium sebife
Oma tieni —
Sebastiania cornuta
Stillingia spinulosa
— 10 changes
Figure 3. Phylogram of one of the equally most parsimonious trees resulting from — ITS analysis, showing branch
lengths within Euphorbieae. Only the species with the least and most changes are
of subtribe Euphorbiinae. Individual а of clades A-D are given in чы жиы (2001). £
5 Wheeler and E. peper-
› share their vegetative
as Euphorbia innocua L.
appear
similarity due to convergence, as they do not group
3
omioides Boiss..
close to the Chamaesyce clade (Figs. 9, 1:
A. Mey.,
1949) to be possibly inter-
Similarly, E. cheirolepis Fisch. & C. sug-
gested by Prokhanof
—
mediate between Chamaesyce and other species of
— is not supported to be closely related
(Figs. 12,
uA these putative relatives are clearly outside
14). At least based on the molecular re-
Chamaesyce, and the characters presented in the
introduction can distinguish all true species of
Chamaesyce. As previously suggested by Dressler
(1957) and Webster (1967), Chamaesyce is derived
from within £. subg. Agaloma (Figs. 12, 14).
Sampling of Chamaesyce was not broad enough
to determine the monophyly of the subsections pro-
posed by Boissier (1862). However, it is noteworthy
that С. acuta and С. angusta, both in Chamaesyce
subsect. Acutae Boiss., lie sister to the remainder
of the species sampled from Chamaesyce subsects.
Gymnadeniae Boiss., Cheloneae Boiss., Chamaesy-
ce, and Hypericifoliae Boiss. Chamaesyce subsect.
Acutae is the only group within Chamaesyce not
possessing the derived C, photosynthetic pathway
otherwise characteristic of the genus (Webster et
al., 1975; Mayfield, 1991).
Al-
though their composition differs between the ITS
Synadenium/Monadenium/Endadenium.
and ndhF analyses, a total of six species from these
genera, here referred to as the Synadenium alli-
ance, were included. They form a well-supported,
monophyletic group (bs 10096 in both analyses, see
10, 11). Both ITS and ndhF also suggest that
Monadenium is paraphyletic with Synadenium and
Figs.
wn for eac һ of the major clades
= Euphorbia.
Endadenium nested inside it. Morphologically, the
genera are united in the possession of fused invo-
lucral glands. They belong to a well-supported
C (bs 78% in ITS, 100% in ndhF) that also
contains Kuphorbia subg. Euphorbia and Lacanthis.
clade
Pedilanthus. As their unusual zygomorphic cy-
athia suggest, the eight species of Pedilanthus in-
cluded here form a well-supported (bs 100% in
both analyses). monophyletic group (Figs. 9. 11).
Based primarily on its involucral appendages and
predominantly Mexican distribution, Dressler
(1957) and Webster (1967) suggested that Pedilan-
thus arose from Euphorbia subg. Agaloma, with E.
fulgens Karw. ex Klotzsch as the closest potential
living intermediate. However, this analysis does not
support any relationship between these two groups.
In fact, they do not even belong to the same major
clade of Euphorbiinae, with E. fulgens (Fig. 13) and
other members of subgenus Agaloma (Figs. 11—13)
belonging to clade D. Beyond being a member of
clade C, the exact position of Pedilanthus is not
strongly supported. In the ITS analysis (Figs. 9, 10),
it lies sister to the clade of Old World taxa that
includes the Synadenium alliance and E. subg. Eu-
phorbia and Lacanthis. However, there is no boot-
strap support for the relationship. The ndhF anal-
ysis (Fig. 11) suggests a relationship with E. elata
in a weakly supported clade (bs 72%) of various
Neotropical species. At least based on biogeogra-
phy. this association is more plausible. Involucral
appendages have arisen on various occasions with-
in Euphorbiinae, and their presence in Pedilanthus
apparently represents an independent derivation of
this feature.
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
E. guerichiana
E. gradyi
е " |elade D
E. pervilleana
Ped. tithymaloides
| clade C
—— cleo,
Anthostema sp. nov.
Anthostema senegalense
Dichostemma glaucescens
Sapium sebiferum
Omalanthus populifolius
Sebastiania cornuta
— 5 changes
E. p
E. — 5
Е.а
Са * ee collinus
i — clade B
| clade A
s
opeplus paucifolius
opatra
Anthostema madagascariense
tillingia spinulosa
Figure 4. Phylogram of one of the equally most -— trees resulting А from E ndhk analysis, showing
branch lengths within Euphorbieae. Only the species with the
c des of subtribe и Individual phylograms of c nsi A—D are given in Ste inmann 1 (2001). E.
^ed, = Pedilanthus
Cubanthus.
as a section of Pedilanthus, but as noted by Mills-
paugh (1913) and confirmed by Dressler (1957),
these two groups do not appear closely related. Al-
Cubanthus was initially described
though Cubanthus was not available for this study,
Euphorbia gymnonota Urb. and E. punicea Sw.
were included (clade C, Figs. 9, 11). Like members
of Cubanthus, these two species possess involucral
glands that are partially situated on the outside wall
of the involucre in contrast to the rim of the invo-
lucre, as is generally the situation in Euphorbia.
Based on this shared feature, an overall similarity
in habit, and an allopatric distribution, Cubanthus
is presumably closely related to these taxa and
probably would associate with them in a molecular
analysis.
Pax (1921) first
proposed this taxon as a section of Euphorbia, but
Euphorbia subg. Trichadenia.
a decade later he synonymized it with Euphorbia
sect. Tithymalus Roeper,
which group within section Tithymalus he believed
without elaborating on
that it belonged. Carter (1985) resurrected E. sect.
Trichadenia Pax as a subgenus and proposed two
Somalica S. Carter (inflorescences
sections: sect.
umbellate and unbranched. with small deciduous
bracts) and sect. Trichadenia (inflorescences
branching. with large persistent bracts). Later she
removed her section Somalica from Euphorbia
subg. Trichadenia and placed it in subgenus Ly-
1988a). Although E. subg. Trichad-
Carter is now generally accepted, its
clopsts (Carter,
ета (Pax) S
exact rank and placement are disputed. Based pre-
the major
= Euphorbia,
ast and most changes are үп for each of
sumably on the fact that many species possess a
pseudoumbellate inflorescence, Gilbert (1987,
1990). for example, preferred to treat it within £.
subg. Esula.
Various species referred to Euphorbia subg. Tri-
chadenia by Pax, Carter, and Gilbert were included
in this study: Euphorbia trichadenia Pax, E. goetzet
Pax, E. platycephala Pax, E. grantii Oliv., and F.
omariana M. G. Gilbert. With the exception of E.
goetzei, which falls out in clade D as related to E.
pirottae N. (Figs. 12, 14).
subgenus Trichadenia are supported to be closely
related and belong to clade A (Figs. 5, 6
Terrac. the members of
). However,
they do not form a monophyletic group. Instead,
they belong to a group together with Euphorbia
Dactylanthes (Haw.) A. Medusea
Meleuphorbia A. Berger,
and Anthacantha (Lem.
sects.
(Haw.) Baill.,
(Haw.) Baill.,
ese sections,
Berger.
Treista
Мм
A. Berger.
which lack a current. subgeneric
placement, all contain tuberculate-stemmed South
African succulents, many of which also possess
glandular involucral processes. Their various char-
acteristics are as follows: E. sect. Anthacantha
(leaf-reduced stems, the inflorescences axillary with
the cyathia borne at the tips of persistent. spiny
peduncles; representative: E. atrispina N. E. Br.):
Dactylanthes highly
the involucral glands with
. sect. (leaf-reduced,
branched, short stems,
long, fingerlike divisions and curved lip at the base:
representative: E. globosa (Haw.) Sims): Ё sect
Treisia (leafy stems and three conspicuous bracts
just below the cyathium; representative: E. clava
Annals of the
Missouri Botanical Garden
E.o
platycephala He. sg. Trichadenia
НЫНЫН
Е monteiri — Pseudeuphorbium
: ME sg. Trichadenia
E. —
E. obesa E. s. Meleuphorbia
E. tubiglans
E. — s. Anmacantha
E. clava — E. s. Tre
E. — оза — —Е. sg Rhizanthi. ium
E. balsamifera ssp. сл — E. sg. Esula
E. meul
E hellii
E. caputmedusae
© L E esculenta RE s. Medusea
”— E. acalyphoides
— = киш ыры
Le sg. Eremophyton
E. — les
E.a
Figure 5.
of the ITS region. Bootstr rap values greater than 50% are shown above the branches. E. = Euphorbia. s. = sectio
8 I 8 I
— subgenus, ssp. — subspecies.
Jacq.); E. sect. Medusea (main stem normally partly
sunken in the ground and with few to many sec-
ondary branches crowning and radiating from the
apex; representatives: E. caputmedusae L. and E.
esculenta Marloth); E. Meleuphorbia (little-
branched to unbranched aboveground stems that
are frequently subglobose and with distinct angles;
representatives: E. tubiglans Marloth ex R. A. Dyer,
f., and E. meloformis Aiton). The
close relationship between the species of subgenus
Trichadenia and these taxa is not surprising and
Gilbert (1987), who noted that
there does not appear to be any important discon-
sect.
E. obesa Hook.
was predicted by
tinuity between these groups. Also closely allied
are E. tuberosa L., the type of E. subg. Rhizan-
thium; E. monteiri Hook. f., the type of E. sect.
Pseudeuphorbium (Pax) A. Berger; and E. lignosa
Marloth, the type of E. subg. Lyciopsis (Boiss.) L.
C. Wheeler sect. Lignosae Pax & K. Hoffm
Euphorbia subg. Rhizanthium. This taxon was
initially proposed by Boissier (1862) as a section
of Euphorbia to accommodate various geophytes
64
— E. s. Denisophorbia
Euphorbiinae clade A, strict consensus of 2160 trees based on a we ighted 1 maximum parsimony analysis
}
yn, sg.
from India and South Africa. Boissier (1862) treat-
ed another geophyte, Euphorbia sessiliflora Roxb.
from Burma (not included here), in his F.
Caulanthium Wheeler (1943) combined
these two sections under the name Rhizanthium
and elevated its rank to subgenus within Euphor-
bia. Gilbert (1987) provided a detailed overview of
the group, concluding that subgenus Rhizanthium,
sect.
Boiss.
as currently circumscribed, is a heterogeneous as-
semblage of many unrelated species. This study
vindicates his contention, and members of this tax-
on belong to various unrelated groups in clades A
and C (see Figs. 5, 6, 10, 11).
A number of species in Euphorbia subg. Rhizan-
thium were included here. Although Euphorbia ses-
siliflora, the sole member of Boissiers Euphorbia
sect. Caulanthium, was not available for inclusion,
the later-described and apparently closely related
E. panchganiensis Blatt. & McCann and E. meenae
5. Carter were represented. Both of these Indian
geophytes are nested within the spine-shield suc-
culents of E. subg. Euphorbia. Euphorbia tuberosa,
trichadenia — E. sg. Trichadenia
E lobosa — E. s. Dactylanthes
гЗ LM E. sg. Lyciopsis s. Lignosae
s da el hala
* nt * E. sg. Trichadenia
E —
E. obesa — E. s. I PN
89 E. clava — E. s. Treisia
99 :. tuberosa Lg sg. Rhizanthium
62 — — Myr E. sg. Lyciopsis s. Somalica
7 E. — ssp. adenensis — E. sg. Esula
p—— E.matabelensis — E. sg. Lyciopsis s. yciopsis
100 bb E. crotonoides
100 [——- E. acalyphoides E. sg. Eremophyton
— — — 4
E. s. Denisophorbia
Figure 6.
of the ndh¥ region. Bootstrap values greater than 50% are shown above the branches. E.
subgenus, ssp. = subspecies
E uphorbiinae clade A, strict consensus of 19,012 trees based on a weighted maximum parsimony —
= Euphorbia, s. = section,
Volume 89, Number 4
2002
Steinmann & Porter
Phytogenetic Relationships in Euphorbieae
the type of subgenus Rhizanthium, is not closely
related to the Indian species, but instead is related
to a group of South African taxa currently recog-
nized as E. subg. Trichadenia (clade A, Figs. 5, 6:
see discussion above). Despite not truly being geo-
phytes, E. longituberculosa Boiss. and E. pirottae
The
—
were placed here by Pax and Hoffmann (1931
first species is not closely related to any of the
above members, but instead its position is near E.
acalyphoides Hochst. ex Boiss. in E. subg. Eremo-
phyton sect. Pseudacalypha Boiss. in clade A (Figs.
5. 6). where it was placed by Carter (1988a). The
second species also shows no affinities with any of
the above members and comes out related to Eu-
phorbia goetzei within clade D (Figs. 12, 14). Fu-
phorbia primulifolia Baker, a Madagascan species
treated in subgenus Rhizanthium by Denis (1921)
and in E. subg. Lacanthis by Gilbert (1987), be-
longs with other Madagascan species currently
placed in subgenus Lacanthis (clade С, Fig. 10).
When
1862),
contained only Euphorbia cuneata Vahl, a spines-
Euphorbia subg. Lyciopsis. first de-
seribed as a section (Boissier, this taxon
cent shrub from Africa with fasciculate leaves. The
section was compared with E. sect. Tirucalli Boiss.
gland characters but distinguished by the fas-
ciculate leaves and non-succulent stems. Pax and
Hoffmann (1931) expanded section Lyciopsis Boiss.
to include three subsections: subsect. Lyciopsis (as
Eulyciopsis). subsect. Espinosae (Pax & K. Hoffm.)
Pax € K. Hoffm., and subsect. Lignosae (Pax & K.
Hoffm.) Pax & K. Hoffm. The latter two were based
on sections proposed a decade earlier (Pax & Hoff-
1921). Wheeler (1943) followed Pax's
in rank
mann in Pax,
circumscription and made changes only
Carter
—
this taxon t
~
when elevating » subgenus.
(1988a) expanded it to include an additional sec-
tion, E. sect. Somalica, a group erected by her to
accommodate about 10 east African species with
woody semisucculent branches, small scarious de-
ciduous bracts, crenulate or pectinate involucral
— and large, often ornamented capsules (Cart-
. 1988b); E. sect. Somalica was previously placed
in E. subg. Trichadenia (Carter, 1985).
Members of all four currently recognized sections
of Euphorbia subg. Lyciopsis were included in this
analysis. The results suggest that the subgenus is
polyphyletic, and its component sections represent
four distinct groups. In both molecular analyses, E.
espinosa Pax and E. guerichiana Pax form a well-
supported lineage (bs 97% in ITS, 100% in ndhF)
in the early diverging portion of clade D (Figs. 12,
14). They are far removed from the remainder of
the sections of subgenus Lyciopsis, all of which be-
long to clade A (Fig. 6). Due to problems obtaining
TS sequences, the other species of this subgenus
were included only in the ndhF analysis. Both А.
lignosa and E. hamata (Haw.) Sweet, members of
subgenus Lyciopsis sect. Lignosae, are placed in a
well-supported clade (bs 87%) together with E. tri-
chadenia and E. globosa. Their relationship here is
not unexpected; Pax and Hoffmann (in Pax, 1921),
when first describing this section, noted that the
appendages of E. lignosa were similar to those of
species in E. sect. Dactylanthes, of which E. glo-
bosa is a memb
The only species of Euphorbia subg. Lyctopsis
sect. Lyciopsis that was included in this study is Æ.
matabelensis. It comes out as an early diverging
member of clade A, sister to E. crotonoides Boiss.
However, this relationship should be viewed with
reservation because there is little morphological
similarity between these two species and no boot-
strap support for this grouping.
The final section, Euphorbia subg. Lyciopsis sect.
— was represented in this study by two ra
cies, E. scheffleri Pax and E.
— two form a well-supported group (bs 99% in
ndhk, Fig. 6
ifera Aiton and various other Euphorbia, Euphorbia
socotrana Balf.
М
in a subclade containing E. balsam-
balsamifera is vegetatively similar to species of
subgenus Lyciopsis sect. Somalica, so the sugges-
tion that they have arisen from common ancestry is
reasonable.
Euphorbia subg. Esula. This subgenus largely
corresponds to Boissiers Euphorbia sect. Tithyma-
lus. Wheeler (1943), when implementing his system
of subgenera, resurrected the application of E.
subg. Esula for this assemblage. Its current use cor-
responds to the circumscriptions of these authors
with some subsequent modifications. For example,
E. ipecacuanhae L., the type of Boissier's section
Tithymalus subsect. Ipecacuanhae Boiss., has been
transferred to E. subg. Agaloma, and Tithymalus
subsect. Inundatae has been proposed to accom-
modate the remainder of the species treated in this
1967). In addition, E. sect. Ti-
Crotonopsideae Boiss. was re-
subsection (Webster,
thymalus subsect.
moved by Radcliffe-Smith (1974) and placed
subgenus Cystidospermum (Prokh.) Prokh. (see dis-
cussion under Euphorbia subg. Eremophyton). As
employed in the discussion below, many of Bois-
sier's subsections are elevated to the rank of section
E. subg. Esu-
when E. sect. Tithymalus is treated as
Si
a
~
With as many as 500 species, this is the largest
subgenus currently recognized within Euphorbia. It
probably is also the most taxonomically difficult.
Annals of the
Missouri Botanical Garden
The greatest diversity is in northern temperate re-
gions. Plants are mostly perennial herbs, but a va-
riety of growth forms exist from diminutive annuals
to shrubs or rarely small trees. The majority of E.
subg. Esula possess alternate lower leaves and a
well-developed pseudoumbellate inflorescence in
which the stem terminates in a whorl of leaves and
a fascicle of three to many branches with opposite
leaves and dichotomous branching; cyathia are sit-
uated in the axils of these upper leaves. The in-
volucral glands lack appendages and are often trun-
cate or bicornute. Exstipulate leaves also
characterize this group, but some members cur-
rently placed here possess stipules, e.g., the spe-
cies of subgenus Esula sect. ia (Raf.) G
L. Webster and section Tithymalus subsect. Inun-
datae.
About 45 species of Euphorbia subg. Esula, from
the majority of Boissier’s subsections, were includ-
ed in this study. Although most of Boissier’s sub-
sections were sampled, not enough members from
each were included to make any definitive infer-
ences concerning their monophyly. Still, a number
of conclusions can be reached.
Euphorbia ipecacuanhae, the basis of Boissier's
section Tithymalus subsect. Ipecacuanhae, was in-
deed supported as being unrelated to other mem-
bers of this subsection (currently treated as E. subg.
Esula sect. Tithymalus subsect. Inundatae) as well
as unrelated to other members of subgenus Esula:
thus, its removal by Webster (1967) is justified. Eu-
phorbia ipecacuanhae belongs with members of
subgenus Agaloma in clade D (see Fig. 13). Also
justified is Radcliffe-Smith’s placement of section
Tithymalus subsect. Crotonopsideae within Euphor-
bia subg. Cystidospermum (see discussion under £.
subg. Eremophyton).
Euphorbia subg. Esula sect. Tithymalus subsect.
Inundatae was represented by three South Ameri-
can species: E. papillosa A. St.-Hil., E. stenophylla
(Klotzsch & Garcke) Boiss., and E. thinophila Phil.
These do not demonstrate any close relationship to
other members of subgenus Esula. Instead, the mo-
lecular evidence supports that they are related to
E. peperomioides Boiss. of section Nummulariopsis
and Е. Phil.
castrum Boiss. Together these form a monophyletic
Boiss. germainit of section Portula-
group in clade C (see Fig. 9).
Euphorbia subg. Esula sect. Balsamis Webb. &
Berthel., a taxon corresponding to Boissier's section
Tithymalus subsect. Pachycladae, is clearly poly-
phyletic (Figs. 5-9, 11). This is not surprising, con-
sidering that the only features uniting its members
are their shrubby habit together with branches that
аге leafy only toward the tips but leafless with
promine nt leaf scars proximally. Euphorbia balsam-
ifera subsp. adenensis (see Figs. 5, 6) does not ap-
pear closely related to any other species of subge-
nus Esula sect. Balsamis sampled here. Instead, it
belongs to clade A and based on the ITS results
(Fig. 5),
Schwartz from Yemen, a vegetatively similar spe-
it has its affinity with E. meuleniana О.
cies. Euphorbia plumerioides Teijsm. ех Hassk.,
also referred to this group by Boissier, shows no
close relationship to E. balsamifera but instead be-
longs to clade С (Fig. 9). Euphorbia plumerioides is
a member of a group of about eight species occur-
ring in Australia, Malesia, and Melanesia (Foster,
1994). In this analysis, the Hawaiian species, E.
haeleeleana D. R. Herbst, is suggested to belong to
this group. Also related is E. boóphthona C. A.
Gardner, an herbaceous plant from Australia pre-
viously treated in subgenus Eremophyton. Both E.
balsamifera and members of the E. plumerioides
group differ from typical members of subgenus Esu-
la by lacking the characteristic pseudoumbellate
inflorescence described above. A third group of
species sampled from subgenus Esula sect. Bal-
samis, E. dendroides L., E. longifolia Lam., E. re-
el., and Ё.
are demonstrated by the molec-
gis-jubae Webb. atropurpurea
Brouss. ex Willd.,
ular evidence to be related to typical members of
the north temperate group of subgenus Esula dis-
cussed below (see Figs. 7, 8), although collectively
these do not form a monophyletic group.
Another group that does not appear related to the
remainder of Euphorbia subg. Esula is its section
Adenorima. This taxon corresponds to Boissier's
section Tithymalus subsect. Laurifoliae; another
synonym is the genus Euphorbiodendron Millsp. As
the last name suggests, the ca. 20 species compos-
ing E. subg. Esula sect. Adenorima are usually
trees. They occur primarily in tropical forests from
Mexico to northern South America and the Carib-
bean and possess the pseudoumbellate inflores-
cence structure typical of subgenus Esula. It is for
this reason that they traditionally have been placed
here. A number of species presently referable to
subgenus Esula sect. Adenorima were included in
E. cestrifolia HBK, E.
laurifolia Juss., E. elata Brandegee, E. tanquahuete
Sessé & Mociño, E. calyculata HBK, and E. gym-
nonota Urb. (Figs. 9, 11). These taxa do not appear
this study: E. punicea Sw.,
closely related to the core Esula group discussed
below, and the two groups are in separate major
C and B,
tively. In addition, the species of subgenus Esula
sect. Adenorima do not form a monophyletic group,
clades of Euphorbiinae, clades respec-
but instead appear to represent an ancestral New
World grade from which various other groups of Eu-
Volume 89, Number 4
Steinmann & Porter
2002 Phylogenetic Relationships in Euphorbieae
80 92 = Schimper Г Е. sg. Tirucalli
E. usambarica
91 E. aphylla
E. regis-jubae
E. atropurpurea
100 E. peplus
100 E. е
Е. —
E. m
E. megalatlantica
E. ro
Е. turczanino
72 = E. amygdaloides
: E. soongari m
7 — — E. esula E. sg. Esula
99 E. disco
lL. E krausiana
E. calyptrata
54 r— Е. lathyris
LL E. myrsinites
100 E. spathulat
— E.al
E. acanthothamnus
E. oblongata
E. depauperata
100 ro Е. pilosa
| L— E. longifolia
E. stricta
Figure 7. Euphorbiinae clade B, strict consensus of 2160 trees based on а weighted maximum parsimony analysis
of the ITS region. Bootstrap values greater than 50% are shown above the branches. E. = Euphorbia, sg. = subgenus.
phorbia have evolved. Dressler (1957) suggested
that this taxon includes the most primitive members
of the
strongly supported here, the species do represent
genus. Although that assumption is not
some of the earliest diverging taxa in clade C and
possess many traits that appear primitive for the
genus.
In both the ITS and ndhF analyses, the remain-
ing species of Euphorbia subg. Esula sampled come
out together (see Figs. 7, 8). This group c orresponds
to the following subsections of Boissiers section Ti-
thymalus: subsect. Decussatae Boiss., subsect. Op-
Carunculares Boiss.,
positifoliae Boiss., subsect.
subsect. Galarrhaei Boiss., subsect. Esulae Boiss.,
and subsect. Myrsiniteae Boiss. They form a well-
supported, monophyletic assemblage (bs 86% in
ndhV) only after the inclusion of the third group of
species discussed above in subgenus Esula sect.
Balsamis and the inclusion of E. schimperi Presl
and A.
been placed in E. subg. Tirucalli, but it should be
mauritanica L. The latter two species have
emphasized that their association with Æ. tirucalli
„ and relatives is based primarily on gross. mor-
phology; the possession of pseudoumbellate inflo-
rescences and the lack of stipules supports their
placement in subgenus sula, as the molecular
data here strongly suggest.
Collectively the taxa mentioned in the previous
paragraph compose clade B and correspond broadly
Г
Euphorbia subg. Esula. The presence of Euphorbia
to the temperate, Northern Hemisphere group o
in temperate regions is almost entirely the result of
the radiation of this lineage, and very few other
Al-
though the great majority of this group are restrict-
species of Euphorbia are found in such areas.
ed to temperate, Northern Hemisphere environ-
ments or high-elevation montane tropical regions,
this
mauritanica and relatives are found in arid tropical
is not a strict characterization. Euphorbia
E
E.
dendroides, E. longifol-
and subtropical regions;
ia, E. regis-jubae, and E. atropurpurea occur in
subtropical vegetation on the Canary Islands; and
E.
tropics.
trichotoma inhabits beaches in the New World
. aphylla
atropurpurea
ue 74 Al ats intica
ped op
esula HE. sg. Esula
myrs unite
anthothamnus |
: Spath ule
. Le meifolia
gure 8.
of us ndhY region. Bootstrap values greater than 50% are shown above the branches. E.
Euphorbiinae clade B, strict consensus of 19,012 trees based on a weighted а — analysis
bg
pilose
= Euphor = subgenus.
468 Annals of the
Missouri Botanical Garden
‚10
96 d tehuacanus
96 Ped. bracteatus
Р, carpus
100 Ped. connatus
7. thinophila — Esula
е qu — БЕ 5. Nummulariopsis
кенер, phyla g. sg. Esula
\ po — E. s. Portulacastrum
ete ‹
ollmeriana E. s. Stachydium
agunillarum ©
51
— (3) E. s. Pteroneurae
г. pteroneura (2
аня (1)
es. — — E. s. Euphorbiastrum
ueri —R. s. оаа
E oia |- E. sg. Esula
unicea
. calyculata E. sg. Esula
Jesi iris — E. sg. Eremophyton
Figure 9. E cp clade С (in part), strict consensus of 2160 trees based on a weighted maximum parsimony
analysis of the ITS ау, values greater than 50% are shown above the branches. E. = Euphorbia, Ped. =
Pedilanthus, s. = s i Em sg. — subgenus, ssp. — subspecies
E. panchganiensis Le. sg. Rhizanthium
. meenae
7. teke
Г. —
E. sg. Euphorbia
1
1
[
E
E
E -qpiphloides
E. abdelkur.
E
і
i г primulifolia
E.
г. perr
.mi H і
. ankarensis I— E. sg. Lacanthis
ii
. capmanambatoensis
. iharanae
E . hedyot oides | А
E. mahabobokensis [—E. s. Denisophorbia
E - elliotii
г. alluaudii —E. s. Deuterocalli
Syn. grantii
fon. elegans Endadenium/
nd. gossweileri Monadenium
fon. magnificum | | Synadenium
— brunellii He. sg. Lacanthis
Figure 10. о clade С (in part), strict consensus of 2160 trees based оп а we id d maximum parsimony
analysis of the ITS region. Bootstrap values greater than 50% are — above the branches. E. = Euphorbia, End. =
Endadenium, Mon. = Monadenium, Syn. = Synadenium, s. = section, sg. = subgenus.
Volume 89, Number 4 Steinmann & Port
2002 Phylogenetic ае in Euphorbieae
. meenae — E. sg. Rhizanthium
. abdelkuri
. drupifera Le sg. Euphorbia
. poissonii
. brunellii
` rubella I- E. sg. Lacanthis
n. grantii | Endadenium/
тїит/
рей *
vides sg. Lacanthis
Дате. — ^s eye
pen — — E. s. Deuterocalli
— E. s. Portulacastrum
{эе hila — E. sg. Esula
:. lactiflua
. calyculata
7. tanguahuete | — E. sg. Esula
г. punicea
* pteroneura (1) — E. s. Pteroneurae
do niana — E. s. Euphorbiastrum
eberbaueri — E. s. Pteroneur
[ | 100 eH sip — E. sg. Esula
"ed. tehuacanus
Ред tithymaloides * Pedilanthus
100 ata — E. sg. Esula
m.» 100 PL 6 ; :lagunillarum Le s. Stachydium
ا поа — Е. s. Pteroneurae
100 г eleele
=== оета | sg. Eremophyton
100 [—— E. perville — E. s. Denisophorbia
ڪڪ | E pisadas EE s. Tirucalli
Figure 11. Euphorbiinae clade С, strict consensus of 19,012 trees based on a weighted maximum —
analysis of the ndhF region. Bootstrap values | greater than 50% are shown above the branches. E. = Euphorbia, End.
= Endadentum, Mon. ian Ped. = Pedilanthus. Syn. = Synadenium, s. = section, sg. = subge
At least from the limited sample examined in this site, asymmetrical leaves and elongate, glandular
study, the primary division within the restricted cir- stipules. These features together with axillary, 4-
cumscription of Euphorbia subg. Esula, discussed glanded involucres are reminiscent of Chamaesyce,
in the previous two paragraphs, is between those to which Wheeler (1943) believed that this section
species that possess tuberculate ovaries and those belonged. Webster (1967) noted that E. peperom-
species whose ovaries are smooth. It remains to be ¿oides bordered Chamaesyce. In this analysis, E. pe-
|
crease in sampling. If so, this would have important is instead related to E. stenophylla, a South Amer-
seen whether this distinction will withstand an in- peromioides shows no affinity with Chamaesyce.
taxonomic implications because many of Boissiers ican member of subgenus Esula sect. Tithymalus
subsections contain both tuberculate- and smooth- subsect. /nundatae (Fig. 9). Although in habit Æ.
ovaried species. peperomioides and E. paranensis differ greatly from
It is worth emphasizing that the occurrence of a оћег members of this subsection, a close relation-
pseudoumbellate inflorescence in all of the major ship is suggested by the common possession of su-
lineages of Euphorbia suggests this is а symple- bulate, sepal-like lobes below the pistillate flowers,
siomorphic feature that was present before the dif- a feature otherwise very rare in subtribe Euphor-
ferentiation of subtribe Euphorbiinae. The struc ture biinae.
bears some resemblance to the inflorescence of : | : .
; Euphorbia sect. Portulacastrum. Two species,
Neoguillauminia and may be homologous. There- . | Dn | 5.
. : PI — Euphorbia germainii and E. pentlandii Boiss. (the
fore, the possession of this type of inflorescence : | : . . à
. — mE latter not included), are contained in this section.
should not be viewed as the defining characteristic ‘a
ban Ы — n . Both are South American annuals with cleft invo-
of E. subg. Esula. Instead, it is the combination of
; lucral appendages. Based on the presence of these
a pseudumbellate inflorescence together with ex-
involucral — посаде, section Portulacastrum has
stipulate leaves that better characterizes the core
been placed in E. subg. Agaloma, e.g., Wheeler
subgenus Esula group, as represented by clade B.
(1943). However, according to the molecular data,
Euphorbia sect. Nummulariopsis. Euphorbia at least E. germainii does not fall out as related to
peperomioides and the very similar E. paranensis the core Agaloma group. Instead, it nests within
Dusén (the latter not included) are the only two South American members of subgenus Esula sect.
species belonging to E. sect. Nummulariopsis. Both Tithymalus subsect. /nundatae in clade С (Fig. 9).
are prostrate Brazilian perennial herbs with oppo- This placement is anomalous, and E. germainii and
Annals of the
Missouri Botanical Garden
E. pentlandii are morphologically incongruous
there because they possess well-developed involu-
cral appendages and lack the characteristic sepal-
like lobes below the pistillate flowers of the sub-
section /nundatae.
Euphorbia sect. Denisophorbia. This is a small
group of approximately 20 species of leafy trees and
shrubs, mostly confined to Madagascar. It was first
proposed as a subsection of Euphorbia sect. Eu-
phorbia by Leandri (1957). Croizat (1972) elevated
the group to the rank of section. As mentioned by
Leandri (1957), section Denisophorbia is difficult to
define. The leaves are entire and alternate to spi-
rally arranged. The cyathia are relatively large, lack
appendages, and are solitary or in terminal pseu-
doumbellate inflorescences. Seven species belong-
ing to this group were included here: Euphorbia
antso Denis, E. denisti гаа E. elliotii Lean-
dri, E. hedyotoides N. . E. mahabobokensis
Rauh, E. pervilleana Baill., and E. tetraptera Baker.
These species do not form a monophyletic clade
but instead come out as representing three separate
groups. Euphorbia antso is the earliest diverging
species in clade A (Figs. 5, 6) and shows no affinity
with the other species of E. sect. pie, edge
sampled here, all of which belong to clade C (Fi
9-11). Euphorbia denisii, E. pervilleana, and E. te-
traptera form a monophyletic group that is sister to
E. tirucalli and relatives, currently treated in sub-
genus Tirucalli. Not closely related to this group
are E. elliotii, E. hedyotoides, and E. mahabobok-
ensis. These form a monophyletic clade sister to
subg. Lacanthis proper.
Euphorbia sect. Denisophorbia was proposed to
be the most primitive group in Euphorbia (Webster
et al., 1982). In part, this may be correct because
Euphorbia antso is the earliest diverging species of
clade A and possesses many of the primitive fea-
tures for the subtribe (see discussion below under
origin and biogeography of Euphorbieae). Also,
with regard to ITS and ndhF molecular a
this species possesses the least amount of genetic
divergence in relation to the outgroup taxa (see
Figs. 3, 4).
Euphorbia subg. Tirucalli. The section Tirucal-
li Boiss. was proposed in Euphorbia to accommo-
date arid-adapted shrubs with long, slender, semi-
succulent branches (Boissier, 1862). The leaves are
reduced and the stem is green and photosynthetic.
The section was elevated to subgenus by Carter
(1985), and she later noted that it contains two
well-defined groups (Carter, 1992a). The first group
ы
corresponds to E. tirucalli and relatives and is
characterized by small scarious bracts, tightly con-
gested inflorescences, and glandular stipules. The
second group corresponds to species such as F.
mauritanica and is characterized by leafy bracts,
pseudoumbellate inflorescences, and a lack of stip-
ules.
Various members of Euphorbia subg. Tirucalli
were included in this analysis, and the two groups
recognized by Carter do not appear closely related.
Instead, their similarities in growth form appear to
have resulted from convergent evolution. Euphorbia
irucalli and relatives (represented here by E. ar-
buscula Balf. f., E. gregaria Marloth, and E. xylo-
phylloides Brongn. ex Lem.) come out in clade C
as sister to a group of leafy shrubs from Madagascar
that are currently treated in section Denisophorbia
(Figs. 10, 11). The remainder of the species of sub-
genus Tirucalli (represented here by E. mauritanica
and E. schimperi) are found nested within the north-
temperate group of subgenus Esula in clade B (Fig.
7). The leafy bracts, pseudumbellate inflorescenc-
es, and lack of stipules of these latter plants sub-
stantiate this placement.
Euphorbia lactiflua Phil. ex Boiss., a shrub from
the deserts of Chile and the only New World spe-
cies referred to this group (Boissier, 1862), is not
related to any other species of the subgenus (Figs.
9, 11). In fact, its initial placement by Boissier is
in itself very peculiar because this species is a
leafy, scarcely succulent shrub. Euphorbia lactiflua
is taxonomically isolated and shows no close rela-
tionship with any other species of Euphorbia. In
this analysis, its affinities are not determined with
precision, but belongs to clade C, in a group of
various Neotropical Euphorbia.
Euphorbia sect. Euphorbiastrum. This taxon
was first erected as a genus by Klotzsch and Garcke
(1860) to accommodate their new species, Euphor-
biastrum hoffmanniana Klotzsch & Garcke. Bois-
sier (1862) then reduced Euphorbiastrum Klotzsch
& Garcke to a section of Euphorbia. Its most dis-
tinctive feature is that the involucres are solitary in
the leaf axils and subtended by a condensed spiral
of small, imbricate bracts. In this analysis, Eu-
phorbia hoffmanniana (Klotzsch € Garcke) Boiss.
occurs in clade C where it is related to E. weber-
baueri Mansf. and E. pteroneura A. Berger in a
well-supported group (bs 91% in ITS, 9896 in
ndhF) together with E. cestrifolia (Figs. 9, 11). Ac-
cording to the ITS evidence (Fig. 9), E. laurifolia
is the basal member of this assemblage, but there
is no bootstrap support for its placement.
Members of this group are morphologically quite
different. Euphorbia cestrifolia and E. hoffmanni-
ana are leafy shrubs; E. weberbaueri is a leaf-re-
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
471
duced, stem-succulent shrub; and E. pteroneura is
a leafy, drought-deciduous herbaceous succulent.
Despite their gross differences in habit, some fea-
tures suggest relationships among these plants. For
both £.
complex,
one, cestrifolia and E. weberbaueri possess
well-developed glandular stipules. As
well, in all taxa the cyathia are closely subtended
and sometimes enclosed in a pair of bracts, the
styles are entire or emarginate, and the seeds are
similar in size and shape and possess a minute car-
uncle. The involucres of E. weberbaueri and E. pter-
oneura possess four glands while those of E. cestri-
folia and E. hoffmanniana possess five glands.
Euphorbia pteroneura was the basis for E. sect.
Pteroneurae A. Berger (1906). Other taxa have been
placed in this section, e.g., E. sipolisii and E. phos-
phorea, but a close relationship between these and
E. pteroneura is not supported here (Figs. 9. 11).
Given the close affinity of E. hoffmanniana and E.
pteroneura, section Pteroneurae is best reduced to
synonymy under E. sect. Euphorbiastrum.
Euphorbia sect. Stachydium. This section is
characterized by a congested, primarily monocha-
sial inflorescence on which the pairs of bracts are
closely imbricate and fold over to obscure the in-
volucres within. There are about five species in
South America and one species, Euphorbia phyl-
(1862).
sect. Stachydium Boiss..
loclada Boiss.. in Namibia. Boissier who
first described £. recog-
nized two subsections: subsect. Americanae Boiss.
(for the American species) and subsect. Capensis
Boiss. (for E. phylloclada).
Five species were in-
cluded here: E. comosa Vell. E. gollmeriana
Klotzsch ex Boiss., E. heterodoxa Müll. Arg.. E.
lagunillarum Croizat, and E. phylloclada. ln this
analysis, the two subsections are placed in different
major clades of Euphorbiinae (clades С and D.
spectively), and a close relationship between them
is not supported. The similarity between the two
subsections is due mostly to the unusual architec-
ture of the inflorescence. Pax (1921) treated Eu-
phorbia phylloclada as a member of E. sect. Pseu-
dacalypha, based presumably on the axillary
cyathia, but there is no affinity between this species
and other members of Pseudacalypha either.
Based on the possession of involucral append-
ages, Euphorbia sect. Stachydium has been sug-
gested to belong to subgenus Agaloma (Wheeler,
1943). However, only E. phylloclada occurs within
this group (see discussion under subg. Agaloma).
The South American species are members of clade
C (Figs. 9, 11) and form a well-supported group (bs
93% in ITS, 100% in ndhF) that is sister to the £.
phosphorea complex, a small assemblage of leafless.
stem-succulents from eastern Brazil. Beyond its as-
sociation with the E. phosphorea complex, the re-
— of section Stachydium to other members
‘the genus is obscure, although the ndhF results
E that it belongs to a weakly supported clade
of various Neotropical taxa, including Pedilanthus
and subgenus Esula sect. Adenorima.
the
most recent modifications in Euphorbia classifica-
tion (Gilbert, 1987; Carter, 1988a). the subgenus
Euphorbia is restricted to about 250 Old World
stem succulents in which the base of each leaf is
Euphorbia subg. Euphorbia. Following
surrounded by a callous pad, termed a spine-shield.
bearing a pair of spiny outgrowths and stipules
modified into spines. It corresponds to Haworth’s
(1812) and Klotzsch and Garcke's (1859, 1860) re-
stricted genus Euphorbia. Boissier treated these
milii Des Moul. (= E.
splendens Boj. ex Hook.) under his E. sect. Dia-
Although Bentham's (1880) and
Pax and Hoffmann's (1931) classifications recog-
~
species together with E.
canthium Boiss.
nized section Diacanthium in the same sense as
Boissier, it was reduced to a subsection of section
Euphorbium Boiss., a taxon these authors used to
accommodate essentially all of the succulent Ku-
Wheeler (1943) followed Ben-
tham’s and Pax and Hoffmann’s circumscription but
Tithymalus:
the name Tithymalus has otherwise usually been
phorbia species.
under the misapplied name E. subg.
applied to the north temperate members of E. subg.
Esula. Webster (1967) essentially followed Wheel-
ers classification but correctly applied the name £.
subg. Euphorbia to this assemblage.
In both the ndhF and ITS analyses, Euphorbia
subg. Euphorbia belongs to a well-supported clade
(bs 78% in ITS, 100% in ndhF) together with sub-
the
11). Within this clade, all of subgenus
genus Lacanthis and Synadenium alliance
(Figs. 10,
Euphorbia sampled form a monophyletic group, but
only after the inclusion of a few additional taxa.
Elaeophorbia, an African group of four species with
drupaceous fruits, has been recognized as a genus
by some (e.g.. Carter,
Webster,
E. drupifera Thonn. and well supported
19884) and a section by oth-
ers (e.g.. 1967). It was represented in this
study by
(bs 100% in both analyses) to be nested within sub-
11).
meenae are also strongly
supported (bs 100% in ITS) to be nested within
subgenus Euphorbia (Figs. 10, 11). Both of these
genus Euphorbia (Figs. 10, In addition, £.
panchganiensis and Е.
species are dwarf geophytes from India that are
very different in appearance from the typical mem-
bers of the subgenus. At least in the mature phase.
thev lack the characteristic spine-shield structure.
472
Annals of the
Missouri Botanical Garden
Euphorbia panchganiensis is apparently closely al-
lied to E. fusiformis Hamilton ex D. Don, which was
placed by Boissier in E. sect. Rhizanthium (= E.
subg. Rhizanthium sensu Wheeler, 1943). Based on
almost identical capsules and seeds, Gilbert (1987)
first suggested that E. fusiformis might have evolved
from Asian members of subgenus Euphorbia. The
molecular evidence supports his suspicions. The
reduction to geophytic herbs has been documented
in at least one African lineage of subgenus Eu-
phorbia (Carter, 1994), and it appears that the In-
dian geophytes represent a parallel derivation of
this growth form from spiny shrubs. The Indian geo-
phytic species should be examined at early onto-
genetic stages to determine if indeed spine-shields
are present in the seedlings.
Carter (1994) recognized two sections and many
subsections in her classification of Euphorbia subg.
Euphorbia, but not enough taxa were included in
this study to test the validity of these groups.
Euphorbia subg. Lacanthis. Lacanthis Raf.
(Rafinesque, 1837) originally contained only La-
canthis splendens Raf. (= Euphorbia тий). The
name slipped into obscurity until Gilbert (1987
resurrected it at the rank of subgenus, applying it
to a group of species from Madagascar previously
—
treated together with the spine-shield taxa of E.
subg. Euphorbia, sensu stricto, discussed above.
Gilbert also included within subgenus Lacanthis
(Raf.)
Baill. ex Boiss. and the
subg. Rhizanthium. As discussed in detail (Gilbert,
1987), there are numerous differences that suggest
these species should be treated separately from the
Gilbert the species of E. sect. Goniostema
Madagascan members of E.
narrowly defined subgenus Euphorbia. For exam-
ple, in subgenus Lacanthis the inflorescences are
much branched (vs. little branched), the bracts are
well developed (vs. greatly reduced), and the seeds
are oblong-cylindrical (vs. ovoid to subglobose).
Additionally, in subgenus Euphorbia the spines are
borne on a differentiated spine-shield and the stip-
ules are represented by prickles just above the leaf,
but in subgenus Lacanthis the spines are actually
the stipules and a spine-shield is absent. Besides
the
subgenus Lacanthis a few geophytes from tropical
Madagascan species, Gilbert also included in
east Africa that are morphologically very similar to
some of those from Madagascar.
A broad array of species from Euphorbia subg.
Lacanthis were sampled.
included E. milit, E. pedilanthoides Denis, E. got-
tlebei Rauh, Е. rossii Rauh & Buchloh, E. primu-
lifolia, E. thouarsiana Baill., E. perrieri Drake, Е.
ankarensis Boiteau, E. millotii Ursch € Leandri,
Those from Madagascar
geroldii Rauh, E. шии Rauh, and E.
tharanae Rauh. From tropical rica, Ё. ru-
ella Pax and E. brunellii Choiv. ex Chiarugi were
represented. All taxa belong to clade C in a sub-
clade containing subgenus Euphorbia and the Syn-
adenium alliance (Figs. 10, 11). The tropical east
African and the Madagascan taxa belong to differ-
ent lineages within this subclade. The Madagascan
members of subgenus Lacanthis are supported as a
monophyletic assemblage belonging to a clade also
containing species not previously associated with
the group. Euphorbia alluaudii Drake, a Madagas-
can endemic very similar in habit to E. tirucalli but
treated by Croizat (1972) in his E. sect. Deuterocalli
Croizat, represents the earliest diverging species
within the group. Another group not previously as-
sociated with subgenus Lacanthis is the E. hedyo-
toides complex of E. sect. Denisophorbia. In this
analysis, the complex forms the sister clade to sub-
genus Lacanthis proper.
It is suggested that the origin of Euphorbia subg.
Lacanthis, sensu Gilbert, is separate from that of
subgenus Euphorbia and spinescence and succu-
lence in these two groups have resulted from in-
dependent derivations.
Euphorbia sect. Arthrothamnus. Klotzsch and
Garcke (1860) first proposed this group as a genus
to accommodate Euphorbia tirucalli and seven spe-
cies from the Cape Region of South Africa. Boissier
1862) later treated it as Euphorbia sect. Arthro-
thamnus. He removed F. tirucalli but expanded the
group to include two species from the West Indies.
—.
The latter two were treated within his Ё s
throthamnus subsect. Americanae Boiss., while the
remainder of the Old World taxa were placed in
section Arthrothamnus subsect. Capenses Boiss.
As discussed further under Euphorbia subg.
Agaloma, the two subsections of Arthrothamnus do
not appear closely related, and this section should
be restricted to about 20 species in South Africa
and Namibia. These are dioecious, dichotomously
branching shrubs with photosynthetic, articulate
branches and small, opposite leaves. Two species
of the group were included here, E. juttae Dinter
and E. rhombifolia Boiss., and the close relation-
ship of both of these is well supported (bs 100%
in both analyses, see Figs. 12, 14). They belong to
clade D and represent an early diverging lineage
of this clade. Further relationships of these species
to other Euphorbia are unclear.
Euphorbia subg. Eremophyton. This group was
first erected by Boissier (1862) as a section of Eu-
phorbia to include Euphorbia eremophila A. Cunn.,
and E.
E. agowensis Hochst. ex Boiss., gueinzii
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
100
82
91
Figure 12.
analysis of the ITS region. Bo
= Chamaesyce, s. = on, sg. = subgenus
A
—
13.
simony ees of the I
section, sg. — subgenus.
Fig. 13
— tyla
E. sg. Poinsettia
E. sg. Agaloma
lobata s. Zygophyllidium
E. sg. Poinsettia
sg. A
E.s
s. ا УТЯИ
Chamaesyce
. polycnemoides
angusta
a
E > —
"Odes
gymnc clada
чапа
Ез е ا
. Sp. nov.
сеа КЕ. sg. Cystidospermum
E. sg. Eremophyton
go ns ee
homi lada Е. s. Stachydium
rombolta E. s. Arthrothamnus
E. sg. — — is
spin s. Es,
* d
. scatorhiza
г. platyclada
`. goetzei — E. sg. Trichaden
sis | sg. — vton
7. Od.
» FOS
SUC
7. inn
а
es
7. aff.
Euphorbiinae clade D (in part), strict consensus of 2
ITS region. Bootstrap values greater than 50% are shown above the branches. E. =
siana
ssiana var. nov
. ceroderma
. antis pue a
aron-rossi
7. ipe
Ж ре ‘olor
7. cassythoides
`. equi Sora
. misella
. macropus
BA di za
. pirottae— E. sg. n
Euphorbiinae clade D (in part), strict consensus of 2160 trees based on a weighted maximum parsimony
strap values greater than 50% are shown above the branches. £.
= Euphorbia, Cham.
xacana
ше — 5
‘ola
E. sg. Agaloma
s. Alectoroctonum/
Cyttarospermum/
— Petalom
E: — pei ula ithymalopsis/
Сой Trichosterigma
j. SEQOVIENSIS
г. leucocephala
a a
осна
ас —
— E. s. Arthrothamnus
unensis
T n
—
E. sg. Agaloma
s. Cyttarospermum/
ichili
Tithymalopsis
2160 trees based on a weighted maximum par-
Euphorbia,
Annals of the
Missouri Botanical Garden
E. gradyi
Е. delicatula
A 3 ое иа E. sg. Agaloma
E. e - s. Alectoroctonum/
67 100 E. sinaloensis vitarospermum/
94 90 E ms — Dichilium/
E. mis Tithymalopsis/
Е. apes ua Trichosterigma
E. sp. no
E. sphaerorhiza
E. oerstedia
E. SÍFIgOsc
99 E. —
— E. pumicola
7 E А H— E. sg. Poinsettia
E. chersonesa 2. Sg. :
E. hormorhiza
E. heterophylla
E. pulcherrima
گے E. Perd dod
E. bilobata E. sg. Agaloma
E. bifurcata
А eriantha s. ZY apko ldu
93 E. jali
99 — Cham. “articulate [hago Р
L— Cham. ¢ sg. Agalor
, appariciana —
e a s. К оа id
100 r— E. glanduligera
lt E —— laa — E. s. Stachydium
— E. sg. Trichadenia
m E. pi — wie a — тугоп
66 29 E. petiolata — E. se. € 'vstidospermum
E. tannensis —E. sg. Eremophyton
35 E Veo ine 2:
- 100 "c
E. scatorhiz E. sg. ——
. polya —
| 100 l E. guerichiana E. sg. Lyciopsis
E. espinosa s. Espinosae
100 — а n i
й olia
— [ge — ibifoli Fe s. Arthrothamnus
Figure 14.
Euphorbiinae clade D, strict consensus of 19,012 trees based on a weighted maximum parsimony
analysis of the ndhF region. Bootstrap values greater than 50% are shown above the branches. E. = Euphorbia, Cham.
= Chamaesyce.
Boiss. It was originally circumscribed to contain
herbs or subshrubs with alternate lower and oppo-
site upper stem leaves; glandular or subulate stip-
ules; axillary or terminal, solitary involucres with
four to five glands lacking appendages; and carun-
culate to ecarunculate seeds. Carter (1985) empha-
sized also petiolate leaves and exserted capsules as
characteristics of subgenus Eremophyton. As these
traits suggest, there is no single synapomorphy that
unites this group, and all of the features used to
delineate the subgenus can be found in various
combinations in other taxa of Euphorbia. In 1880,
Bentham synonymized Boissier's E. sects. Bongium
Boiss., Cheirolepidium Boiss., and Pseudacalypha
under his expanded section Eremophyton, and this
system was followed by Wheeler (1943) when he
elevated the section to subgenus. Euphorbia sect.
м
Cheirolepidium was removed by Prokhanov (1933
and formed the basis of his genus Cystidopermum
Prokh. Cystidospermum was later reduced to a sub-
genus of Euphorbia (Prokhanov, 1949), and Rad-
cliffe-Smith (1974) agreed that subgenus Cystidos-
sufficiently distinct from the
permum was
remainder of E. subg. Eremophyton to warrant its
separate recognition. He additionally referred to it
Boissiers E. sect. Tithymalus subsect. Crotonopsi-
deae.
Two species of Euphorbia subg. Cystidospermum
(sensu Radcliffe-Smith, 1974) were included in this
study: E. cheirolepis and E. petiolata Banks & Sol.
Various species of subgenus Eremophyton proper
were also included: E. boophthona, E. pirottae, E.
agowensis, E. scatorhiza S. Carter, E. cheirolepis, E.
eremophila, E. tannensis Spreng., E. polyantha Pax,
phoides. In the analyses these taxa do not group
crotonoides, E. longituberculosa, and E. acaly-
together and are widely scattered throughout the
6, 9, 11, 14).
Therefore, subgenus Eremophyton, as currently rec-
subtribe Euphorbiinae (Figs. 5,
ognized, is highly polyphyletic.
three species of Euphorbia subg. Eremophy-
ton sect. Pseudacalypha included in this study (£.
acalyphoides, E. crotonoides, and E. longitubercu-
losa) all grouped as closely related members of
clade
monophyletic group but instead a grade of taxa
A. However, they did not form a strictly
(Figs. 5, 6). Interestingly, E. matabelensis (a mem-
ber of subg. Lyciopsis sect. Lyciopsis) also occurred
together with the species of subgenus Eremophyton
sect. Pseudacalypha as sister to Euphorbia croto-
Volume 89, Number 4
2002
Steinmann & Porter 475
Phylogenetic Relationships in Euphorbieae
noides in the ndhF analysis, although there is no
bootstrap support for this relationship and the two
are morphologically very different.
One of the more unusual results of this study is
the placement of Euphorbia boóphthona, an Aus-
tralian member of subgenus Eremophyton, in a
94% in ITS) together
with E. plumerioides and E. haeleeleana, two ar-
strongly supported clade (bs
borescent Pacific Island taxa presently treated in
subgenus Esula sect. Balsamis. In most respects,
E. bodphthona closely resembles other Australian
members of subgenus Eremophyton, e.g.. E. tan-
nensis and E. eremophila, and these Australian taxa
have been thought to represent a natural group
(Hassall, 1977). It is worth noting that according &
Hassall (1977), E.
two species as well as the
>
boóphthona differs from these
other Australian mem-
bers of the subgenus in being an octoploid (n =
28), but no tetraploid has yet been found in this
subgenus.
Radcliffe-Smith's (1974) decision to unite Bois-
siers Euphorbia sect. Tithymalus subsect. Croton-
opsideae (represented here by E. petiolata) with
subgenus Cystidospermum (represented here by А.
cheirolepis) was supported, and the two species are
sister to each other (bs 92% in ITS, see Fig. 12) in
this analysis. They belong to a clade also contain-
ing two members of E. subg. Eremophyton proper,
E. tannensis and E. eremophila. Thus, subgenus
Cystidospermum probably is best treated as a syn-
onym of subgenus Eremophyton. The Madagascan
endemic E. platyclada Rauh, whose relationship
with other Euphorbia was previously unknown
(Rauh, 1998), also groups with these taxa in the
ndhV analysis, but this relationship is not well sup-
ported (Fig. 14). In the ITS analysis (Fig. 12). F.
platyclada comes out with E.
As suggested by Carter (1992b), these
latter two species are sister taxa in so far as this
scatorhiza and Е.
polyantha.
sample is concerned.
Many species of Euphorbia subg. Eremophyton
possess a great similarity to those of subgenus Aga-
loma, especially taxa of its section Zygophyllidium
Boiss. For example, contrary to reports in the lit-
erature, the southwest Asian E. petiolata and E.
cheirolepis actually possess involucral appendages.
Undoubtedly, if these species occurred in the New
World, they would be treated within subgenus Aga-
loma without question. Therefore, the molecular ev-
idence that E. subg. Agaloma has either evolved
from a portion of subgenus Eremophyton or that the
two are sister taxa and share common ancestry is
plausible.
Euphorbia subg. Agaloma. This taxon is based
on another one of Rafinesque’s genera, Agaloma
af. It was first erected (Rafinesque, 1838) to ac-
commodate Euphorbia corollata L. and two related
species. Interestingly, Rafinesque also published
six other genera that correspond to the current cir-
cumscription of this subgenus (Aklema Raf.. Lepa-
dena Raf., Peccana Raf., Petaloma Raf.,
Raf., and Zalitea Raf.). Euphorbia subg. Agaloma
4). At
was first treated at this rank by House (192
Vallaris
that time it only accommodated E. corollata L., and
other members of modern subgenus Agaloma were
placed in subgenus Lepadena (Raf.) House. Wheel-
er (1943) was the first to adopt subgenus Agaloma
in its current circumscription, a concept corre-
sponding to E. sect. Adenopetalum Boiss., sensu
Bentham (1880) and Pax and Hoffmann (1931).
The group is broadly defined to contain New World
members of Euphorbiinae with petaloid involucral
appendages but excludes species of Chamaesyce
and Pedilanthus. Beside the presence of involucral
appendages, little else unites all members of the
group. Trees, shrubs, perennial herbs, geophytes,
annuals, and stem-succulents are represented. The
subgenus comprises about 150 species and is con-
difficult 1975;
). In addition, it has frequently
sidered taxonomically (Johnston,
Buck & Huft, 1977
been suspected of being paraphyletic,
Chamaesyce and E. subg. Poinsettia nested within
with both
=
Euphorbia subg. Ава ота was the best-sampled
group in this study. Fifty-six species from all of its
recognized sections were included. With the excep-
tion of two of these sections, Euphorbia sects.
Stachydium and Portulacastrum (discussed previ-
ously), all species of subgenus Agaloma belong to
12-14).
Indeed, previous suspicions were supported, and
clade D and form a single subclade (Figs.
both Chamaesyce and E. subg. Poinsettia are shown
to have evolved from within subgenus Agaloma.
Therefore, a monophyletic subgenus Agaloma must
Poin-
also include both Chamaesyce and E. subg.
settia. This entire subclade is hereafter referred to
s the Agaloma alliance.
The r
that Euphorbia phylloclada and E. glanduligera
Рах, two annual African species from the Namibian
analysis strongly supports (bs 93%)
desert, also belong to the Agaloma alliance (Fig.
14). Both species possess well-developed involucral
appendages and morphologically are easily accom-
modated within subgenus Agaloma, where they cer-
tainly would have been placed if it were not for
their African distribution. Boissier treated Кирһог-
bia phylloclada as the sole representative of E. sect
Stachydium subsect. Capensis. However, the molec-
ular data do not support a close relationship be-
Annals of the
Missouri Botanical Garden
tween this species and other members of the section
Stachydium). Pax
(1894b) placed E. glanduligera in Chamaesyce,
(see discussion under E. sect.
and the similarity is indeed strong. In the ITS anal-
ysis, these two species are placed in a clade sister
to some members of subgenus Eremophyton (Fig.
12). However, this relationship lacks any bootstrap
support, and the ndhF analysis probably provides
a more accurate reflection of relationships.
Within the New World,
group in the Agaloma alliance consists of Euphor-
the earliest. diverging
bia subg. Agaloma sects. Ephedropeplus Müll. Arg.
and Crossadenia Boiss. The former section is rep-
pep here by E. appariciana Rizzini, E. sp. nov.
. and E. gymnoclada Boiss., while the latter sec-
tion is represented here by E. crossadenia, E. ses-
silifolia, and E. sarcodes. In the ITS analysis, these
two sections form a single well-supported clade (bs
99%) sister to the remainder of the Agaloma alli-
12, 13). The sample in the ndhF study
was not sufficient to investigate their monophyly,
ance (Figs.
but here too the single species included, E. appar-
iciana, is sister to the remainder of the Agaloma
alliance. Both sections are restricted to eastern Bra-
zil and represent a morphologically diverse assem-
blage of about a dozen arid-adapted perennial
The in-
volucral appendages are deeply cleft to subentire,
herbs, small shrubs, or stem succulents.
and in one species they are lacking. Traditionally
subgenus Agaloma sects. Crossadenia and Ephed-
ropeplus have been separated by the possession of
opposite or whorled, highly reduced scalelike
leaves in section Ephedropeplus and the possession
of alternate, well-developed leaves in section Cros-
sadenia. However, this distinction does not appear
to hold, and with respect to each other, the sections
are not monophyletic. Therefore, E. subg. Agaloma
sect. Ephedropeplus is best reduced to synonymy
under section Crossadenia.
In both the ITS and ndhF analyses, Chamaesyce
is the next diverging lineage in the Agaloma alli-
ance, and it is sister to the remainder of the spe-
cies, excluding the earlier diverging Namibian taxa
and members of Euphorbia subg. Agaloma sect.
Crossadenia (Figs. 12-14). However, statistical
support for this topology is lacking. What is sup-
ported is that Chamaesyce is an ancient lineage that
diverged earlier in the evolution of the Agaloma
alliance. Beyond this, however, the exact relation-
ship of its members to other Euphorbiinae is not
evident.
After the divergence of Chamaesyce, there are
two major groups within the Agaloma alliance.
These are sister to each other and present in both
the ITS and ndhF analyses, with considerable sup-
port (bs 99%) in the latter (Figs. 12-14). The first
clade contains species currently placed in Euphor-
bia subg. Agaloma sect. Zygophyllidium together
with members of subgenus Poinsettia; the second
clade contains the remainder of sections treated in
subgenus Agaloma.
Euphorbia subg. Agaloma sect. Zygophyllidium
corresponds to a North American and Mexican as-
semblage of about a dozen species. The group is
poorly defined, but united by their annual or rarely
short-lived perennial habit. In addition, many spe-
cies possess serrate leaves. The section was rep-
resented in this study by £. idoli Engelm., E.
bilobata Engelm., Е. eriantha Benth., Е. exstipulata
Engelm., E. jaliscensis B. L. Rob. & (жеши, and
E. lacera Boiss. It is not monophyletic because sub-
genus Poinsettia is nested within it (see discussion
under E. subg. Poinsettia).
Sister to the Zygophyllidium/Poinsettia group is
the core of Euphorbia subg. Agaloma, an assem-
blage comprised of section Alectoroctonum
(Schltdl.) Baill.
Benth.,
section Cyttarospermum Boiss. (representatives: E.
E. aff. ariensis HBK, E. calcicola
delicatula Boiss., E. eglandulosa V. W.
E. gradyi V. W. Steinm. & A. Ram.-Roa,
E. graminea Jacq.,
Steyerm., E. lagunensis Huft, E. misella S. Watson,
Rob. & Greenm.,
(representatives: E. colletioides
E. leucocephala Lotsy, and E. sp. nov. 2),
adiantoides Lam.,
Fern., E.
Steinm.,
E. guatemalensis Standl. &
E. oaxacana B. L. E. ocymoidea
L., Е. rzedowskii Mc Vaugh, E. segoviensis (Klotzsch
& Garcke) Boiss., E.
orae E E E. subpeltata S. — E. succedanea
sinaloensis Brandegee, E. son-
L. С. Wheeler, and E. whitei L. C. Wheeler), sec-
tion A Boiss. үжен эч E. insulana
Vell. and E. oerstediana (Klotzsch € Garcke)
Boiss.), section Petaloma (Raf.) Boiss. (represen-
tative: E. bicolor Engelm. & A. Gray), section Ti-
thymalopsis (Klotzsch & Garcke) Boiss. (represen-
tatives: E. A. Н. Holmgren € N. H.
Holmgren, E. innocua L. C. Wheeler, E.
L., E. macropus (Klotzsch € Garcke)
Boiss., and E. sphaerorhiza Benth.), and section Tri-
aaron-rossil
cuanhae
chosterigma (Klotzsch & Garcke) Boiss. (represen-
tatives: E. antisyphilitica Zucc., E. californica
Benth., E. ceroderma 1. M. Johnst., E. fulgens Karw.
ex Klotzsch, E. gentryi V. W. Steinm. & T. F. Daniel,
E. macvaughii Carvajal & Lomelí, E. misera
Benth., E. rossiana Pax, and E. rossiana var. nov.).
In comparison to members of subgenus Agaloma
sect. Zygophyllidium, species of this group are
highly variable in habit, and the leaves are strictly
entire. This group accounts for nearly 90% of the
species that have been treated in subgenus Aga-
loma, and it ranges from Argentina and Chile to
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
the northern United States and the Caribbean. Col-
lectively, these sections form a monophyletic group
in both analyses. However, the clade only has boot-
strap support (94%) in the ndhF analysis (Fig. 14).
Here too belong E. cassythoides Boiss., a Caribbean
taxon previously placed in Euphorbia sect. Arthro-
1862) and E. equisetiformis A.
yalapagos Islands
thamnus (Boissier,
Stewart, an endemic to t
whose relationship to other Euphorbia was uncer-
1971).
Although the sample of species is not broad
tain (Burch,
enough to make a definitive conclusion, this study
supports Parks (1998) narrow circumscription «
Euphorbia subg. Agaloma sect. Tithymalopsis. Ac-
cording to the molecular data, the section, as de-
fined by Huft (1979) and Boissier (1862), is poly-
phyletic and composed of at least three separate
neither Huft nor Boissier can
lineages. However,
be criticized. In fact, the placement of similar-ap-
pearing, geophytic taxa such as E. macropus and
E. sphaerorhiza in separate clades with nongeophy-
tic taxa is odd and represents one instance where
Of the re-
maining five sections of subgenus Agaloma, only
molecular results are counterintuitive.
sections Petaloma and Dichilium are likely to be
monophyletic, but not enough taxa from these were
included to test this hypothesis. What is clearly
shown is that subgenus Agaloma sects. Alectoroc-
tonum, Cyttarospermum, and Trichosterigma are
polvphvletic.
Euphorbia subg. Poinsettia. According to the
most recent treatment of Euphorbia subg. Poinsettia
(Mayfield, 1997), this taxon is a strictly New World
assemblage of 24 species. Dressler (1962) recog-
nized 11 species. The group is characterized by
cup-shaped involucral glands that are generally re-
duced to one (rarely more) per involucre. Pigmen-
tation of the subeyathial leaves, as exemplified by
E. pulcherrima. occurs in many but not all of the
Growth form varies from annuals,
species, peren-
nial herbs, geophytes, to shrubs. Nine species of
subgenus Poinsettia were included in this study: E.
chersonesa Hult, E. heterophylla L.. E. hormorhiza
Radcl.-5m.,
ma, E. pumicola Huft, E. radians Benth.. Е.
E. pentadactyla Griseb., Е. зүн!
stri-
and E. zonosperma ar Arg.
gosa Hook. & Arn.,
In both analyses, these species are nested wills
a clade that corresponds to Euphorbia subg. Aga-
loma sect. Zygophyllidium (Figs. 12, 14). This is
not surprising because some species of this section
(e.g.. E. bifurcata) possess involucres with a single
gland and nearly identical seeds to those of mem-
bers of subgenus Poinsettia. In addition, other taxa
(e.g.. E. jaliscensis) possess panduriform leaves that
í
are otherwise known only from a few species of
subgenus Poinsettia. The subgenus is noteworthy
because of the extreme amount of genetic diver-
gence among its members. It forms a monophyletic
group only in the ndhF analysis (Fig. 14). In the
ITS analysis, subgenus Poinsettia comes out in two
groups nested within subgenus Agaloma sect. Zyg-
ophyllidium (Fig. 12). A close affinity between sub-
genus Poinsettia and species of subgenus Agaloma
1962;
Webster, 1967), but the two groups do not appear
closely related (Figs. 12-14). Dressler's (1962) re-
moval of E. eriantha from subgenus Poinsettia
sect. Dichilium has been suggested (Dressler,
^
justified by molecular evidence. Euphorbia cherso-
E. heterophylla L.
was placed in subgenus Poinsettia by Millspaugh
nesa (= var. eriocarpa Millsp.)
(1889) but placed in subgenus Agaloma by Hufi
(1984). The molecular data strongly support that it
is sister to E. pumicola in subgenus Poinsettia.
Morphologically, E. chersonesa is noteworthy be-
cause it is intermediate between subgenus Agaloma
sect. Zygophyllidium and other members of sub-
genus Poinsettia in involucral gland characteristics;
the glands are not deeply cupped, but they do ap-
parently lack appendages.
The herbaceous habit of the species of Euphorbia
subg. Agaloma sect. Zygophyllidium and many of
the early diverging members of subgenus Poinsettia
suggests that woodiness in species such as £F. pul-
cherrima is secondarily derived. In addition, the
lack of involucral appendages appears to represent
a reversal to unappendaged involucres.
ORIGIN AND BIOGEOGRAPHY OF EUPHORBIEAE
The tribe Euphorbieae demonstrates a complex
The earliest diverging
African
The next diverging
biogeographical pattern.
subtribe Anthosteminae, is strictly
1, 2).
clade. subtribe Neoguillauminiinae, occurs in Aus-
tralia and New Caledonia (Figs. 1, 2). Of the four
major clades of Euphorbiinae, one, clade A (Figs.
clade.
and Madagascan (Figs.
э. б). is stric ‘tly African and Madagascan, and one,
clade B (Figs. 7. 8). is primarily distributed in tem-
perate regions of the Northern Hemisphere. The
(Figs. 9-11) and D (Figs.
12-14), are widespread, but the earliest diverging
Africa
and Madagascar. If present-day distributions are in-
two remaining clades, €
lineages within these two clades occur in
dicative of historical ranges, then the molecular ev-
idence is consistent with a hypothesized African
origin of the tribe before the breakup of Gondwan-
aland.
Although clade B is almost entirely Laurasian in
distribution, there are a few members in Africa. The
478
Annals of the
Missouri Botanical Garden
clade is absent from Australia and essentially ab-
sent from South America, where only two species
occur, و spathulata and E. philippiana
Boiss. The first of these has an amphitropical dis-
tribution also occurs in North America. The
second is a Chilean endemic similar to North
American species. Considering that approximately
85% of the temperate South American species of
angiosperms have an origin in the Northern Hemi-
sphere (Raven, 1963), the presence of these species
in South America is likely the result of dispersal
Within clade B there
are two major subclades, and both possess species
events from North America.
in North America and Eurasia. It is possible that
the current distribution of clade B represents either
an ancient dispersal event to Laurasia or evidence
that the diversification of Euphorbiinae predates
the split up of Pangaea.
Clade C has a wide distribution that is best de-
scribed as pantropical. According to the ndhF re-
sults herein, the earliest diverging lineages of this
clade occur in Africa, Madagascar, and Australasia.
The Hawaiian endemic species Euphorbia haelee-
leana belongs here and appears to have arrived at
the islands by long-distance dispersal from related
taxa, e.g., Euphorbia plumerioides, in the Pacific
Islands. In the ndhF analysis, all of the Neotropical
members of clade C form a single lineage that is
sister to a lineage containing the subgenera Eu-
phorbia and Lacanthis together with the Synaden-
ium alliance, again suggesting that these two groups
arose before the breakup of Gondwanaland. Simi-
larly, the sister clade relationship of subgenus La-
canthis and subgenus Euphorbia together with the
Synadenium alliance suggest that the common an-
cestor of both these groups was present before Mad-
agascar began to separate from Africa.
Clade D has a distribution similar to clade С.
Both ITS and ndhF analyses indicate that the ear-
liest diverging lineages are in Africa, Madagascar,
Also like in clade C, all of the
New World species belong to a single lineage. In-
and southwest Asia.
terestingly, according to the ndhF evidence, of the
two earliest-diverging groups in the lineage con-
taining the New World taxa of clade D, one occurs
in Namibia and the other occurs in arid eastern
Brazil. Thus, there is a clear, well-supported link
between New and Old World species in this group.
The near parallel distribution in the sister clades
C and D provides further evidence that the distri-
bution of New and Old World taxa is the result of
vicariance caused by the breakup of Gondwana-
land. Therefore, the molecular evidence corrobo-
rates that despite being a very specialized group,
the Euphorbiinae are also a very ancient group.
Based on biogeographic patterns, Croizat (1940)
postulated that Euphorbia was already differenti-
ated into modern subgenera by the mid Cretaceous
(ca. 100 million years ago), and this appears quite
possible
The fruits in Euphorbieae are generally dry, ex-
plosively dehiscent, and initially dispersed only a
few meters from their source. For this reason, there
are few instances of long-distance dispersal within
the tribe. One notable exception involves species
of Chamaesyce, a primarily New World taxon whose
range parallels that of the entire Euphorbieae. Web-
ster (1967) proposed that this group probably orig-
inated in the New World and molecular evidence
supports his hypothesis. Because many Chamae-
syce possess mucilaginous seeds, much of their
great success in distribution is likely because seeds
are able to adhere to animals and thus achieve
long-distance dispersal (see discussion in Jordan &
Hayden, 1992).
Euphorbia and helps explain why Chamaesyce is
This feature is otherwise rare in
one of the few relatively derived taxa within Eu-
phorbiinae that exhibits a transoceanic distribution.
Webster et al. (1982) stated that the most-likely
primitive Euphorbiinae were Euphorbia sect.
Denisophorbia. Dressler (1957) believed that they
Al-
though the molecular evidence does not support
were in sections Balsamis and Adenorima.
their hypotheses, it does not provide solid insight
as to what actually is the most primitive Euphorbia
either, and it appears difficult to identify a single
most-ancestral group. Dresslers and Webster's
choices are sound because all three sections rep-
resent early diverging lineages within the genus.
The molecular evidence does provide some in-
sight as to the features that ancestral Euphorbia
likely possessed. First. they were probably a trop-
ical tree or shrub, because early diverging lineages
within the various clades of Euphorbia are mostly
woody tropical plants. Phyllotaxy was presumably
spiral, and it is likely that a pseudoumbellate in-
florescence structure was well developed: these
characteristics are found in all the major clades of
Euphorbia, and their widespread presence is par-
simoniously explained by their presence in the
common ancestor of all four major clades. Stipules
were probably absent or minute and glanduliform.
There were likely five glands on the involucre. Al-
though gland reduction is common, in many species
the
first involucre formed in an inflorescence often pos-
with reduced glands, e.g., E. graminea Jacq.,
sesses five glands and only subsequent involucres
possess fewer glands. Presumably involucral ap-
pendages were absent, because they are present
only in derived groups of Euphorbiinae. It is note-
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
worthy that the Madagascan endemic Euphorbia
antso is the least derived species with regard to
molecular evolution (see Figs. 3, 4) and possesses
most of the features discussed above as ancestral.
Although the ancestral condition is probably woody,
in terms of species number, the two most successful
lineages within the subtribe, Chamaesyce and the
temperate E. subg. Esula clade, are primarily com-
prised of herbaceous species.
The large amount of diversification within Eu-
phorbiinae compared to the other subtribes of Eu-
phorbieae and the sister tribe Hippomaneae sug-
gests that one or more key innovations may have
promoted rapid evolution and diversification within
this lineage. If such innovations actually exist, they
likely involve the eyathium. Although cyathiate in-
florescences also occur in Anthosteminae and Neo-
guillauminiinae, both these subtribes contain very
few species and have narrow distributions. The par-
ticular feature that occurs in. Euphorbiinae but is
absent in the other subtribes is that, with few ex-
ceptions, the nectar-producing involucral glands
are situated on the rim of the involucre and not
partially enclosed within this structure. Therefore,
this feature may have greatly enhanced successful
insect attraction and pollination, thus giving mem-
bers of Euphorbiinae a selective advantage, which
in turn has driven the patterns of speciation and
diversification presently observed in extant mem-
bers of the subtribe.
COMPARISON WITH MORPHOLOGICAL DATA
There are two previous phylogenetic studies of
the Euphorbieae based on morphological characters
(Park, 1996; Park & Elisens, 2000). The first. of
these treated only the New World taxa of subtribe
Euphorbiinae, while the second treated the entire
tribe. The molecular results differ drastically from
the results obtained in either analysis. Even the two
separate morphological analyses yielded different
topologies, and in neither study was there signifi-
cant statistical support for the majority of the
clades. Part of the problem with these morpholog-
ical analyses may be that too few characters were
included to resolve the taxa. For example, in the
first study of the New World species only 37 char-
acters were used to resolve relationships among 49
terminal taxa, in this case species groups. In ad-
dition, many characters employed in both analyses
are highly variable even among closely related spe-
cies, and their use to discern relationships within
such a large and diverse tribe as Euphorbieae is
unlikely to provide accurate results.
FUTURE OF EUPHORBIINAE CLASSIFICATION
The current classification of Euphorbieae sub-
s
tribes Anthosteminae and Neoguillauminiinae
well supported herein and recognizes only mono-
phyletic groups. No taxonomic adjustments are nec-
essary for these subtribes, but as mentioned above,
problems remain for the classification of subtribe
К uphorbiinge.
‘he current taxonomic trend is going in the di-
rection of splitting Euphorbia. This has already
been observed with Websters (1967) removal of
More Carter. (1994: 378)
stated that subgenus Euphorbia “could be separat-
Gilbert
235) also leaned toward an eventual dis-
Chamaesyce. recently,
ed as a genus in its own right." Likewise,
(1987:
memberment of Euphorbia, but well aware of the
"profound nomenclatural consequences? associated
with such a change, he justified only changes in
— within Euphorbia.
Ve disagree that Euphorbia should be divided.
* opinion is that the best long-term solution to
the problem of Euphorbiinae classification is to ex-
pand Euphorbia to encompass all members of the
subtribe. Some might contend that this is an un-
desirable step backward 250 years to Linnaeus's
broad concept of Euphorbia. However, we believe
that this solution is more favorable than leaving the
genus in its current paraphyletic circumscription or
restricting Euphorbia to only the subgenus Kuphor-
bia.
The first taxonomic problem with limiting Fu-
phorbia to subgenus Euphorbia is that ca. 90% of
the species currently in the genus would need to
be accommodated in other genera. Thus, Euphor-
bia, a well-known and easily identifiable taxon
known throughout the world, would no longer exist
in most parts of the globe. Instead, there would be
a multitude of genera completely unknown to most,
and the boundaries and circumscription of these
would be vague and certainly debated for quite
some time. On the contrary, a broad Euphorbia
would require changing the names of ca. 100 (vs
^
а. 1700!) species and would only affect groups
with relatively limited distributions. These changes
would only mildly broaden the current concept of
Euphorbia to encompass taxa that possess an un-
usual involucral morphology, a feature we believe
has received undue taxonomic weight.
Another reason that splitting Euphorbia is un-
satisfactory concerns the unusual nature of evolu-
The situation here is that
—
ion within this group.
basic cyathial morphology in the genus is highly
but morphology is highly
conserved, vegetative
à
plastic. This has led to much parallel evolution in
Annals
cour — Garden
growth form with little change in floral form. There-
fore, if the relationships suggested by the molecular
evidence do indeed accurately depict the phylog-
eny of the group, then there are a number of well-
defined monophyletic lineages nested within a par-
aphyletic background of relatively undifferentiated
groups or groups that have undergone a high degree
of parallel vegetative evolution.
To propose various genera whose members are
superficially nearly identical fails to serve one of
the primary purposes of a system of classification,
e., to provide a predictable system that allows for
the separation of taxa and for the ability to make
assumptions about relationship based on morpho-
logical features. If Euphorbia classification is to ac-
tually reflect relationship, then there will be excep-
tions and inconsistencies in defining new genera.
We fear that any system that attempts to dismember
Euphorbia will continue to have only limited suc-
cess and acceptance.
Croizat (1965: 574) emphasized the problem as-
sociated with subgeneric groups within Euphorbia
and stated "the infraspecific combinations of char-
acters are so intricate as to make it really difficult
to identify a truly natural subgeneric taxon." The
same problem occurs with an elevation in rank to
genus, but the broad implications are greatly more
severe. Recognizing highly similar subgeneric taxa
is much less of a problem because such rank is
usually of primary interest to specialists in the
group. Webster (1967: 398) stated, "If the various
microgenera of Euphorbieae cannot be easily dis-
tinguished, there seems little reason to adopt them
simply because they represent evolutionary units;
it is quite easy to discuss the evolution of these
taxa if they are referred to as infrageneric compo-
nents of Euphorbia.” We could not agree more.
As mentioned in the introduction, the majority
of genera currently segregated from Euphorbia are
distinguished by cyathial traits. In. some regards,
cyathial morphology has received undue weight,
and this may be because there has been great di-
versification in vegetative morphology throughout
Euphorbiinae, but the overall structure of the cy-
athium has remained relatively conserved. Thus,
changes in cyathial morphology are viewed as tax-
onomically significant occurrences, and little sys-
tematic emphasis has been placed on other fea-
tures, including relationships. Leach (1973: 32),
when describing Endadenium, justified his new ge-
nus on the basis of its distinctive arrangement of
glands within the involucre, further saying that En-
"combines so many of the characteristies
African members of the tribe that its
dadenium
of the other
assignment to any one of the genera involved would
so effectively blur, if not nullify, the diagnostic cri-
teria as to make it virtually impossible to retain
these as separate genera on any but purely arbitrary
grounds." However, the current recognition of gen-
era separated solely on the basis of cyathial char-
acters while ignoring, as in the case of Endaden-
ium, all of the characters that do indeed blend the
segregate genera into Euphorbia, is in itself some-
what arbitrary.
One of the arguments frequently used to justify
the division of Euphorbia is that the genus is so
large and heterogeneous that it is unwieldy. The
incredible speciation and diversification that have
occurred within Euphorbia are largely biological
facts. Dismembering the genus would make it
smaller and less heterogeneous but collectively as
unwieldy an assemblage as ever. Another argument
advocated by splitters is that the generic concept
used for these assemblages is too broad in com-
parison with other groups; for the sake of taxonomic
consistency, Euphorbia must be divided. In re-
sponse to this, it is worth pointing out that no two
of plants are completely comparable
Each has an independent evolutionary
the argument that Euphorbia
lineages
each other.
history. Therefore,
should be divided simply because its evolutionary
past has involved amazing speciation, dispersal,
and diversification is unfounded. In our opinion,
recognizing the genus in its broad sense has some
benefit because it conveys the incredible evolution-
ary history of this group.
Can a workable system of classification be de-
veloped for the Euphorbiinae that is based on the
concept of monophyly? We believe that the answer
is yes, but the units of such a classification should
be infrageneric. Certainly many modifications will
be necessary and time required before a stable sys-
tem of classification is developed. However, it is
better that such a process be undertaken at the
infrageneric level, where the far-reaching nomen-
clatural implications associated with generic
changes can be avoided.
Literature Cited
pee Euphorbiaceae Neo-Caledonicae. Ad-
Baillon, H.
D
=
5
М
1992, Ее netic utility of the internal
spacers of nuclear ribosomal DNA in
plants: b exile the Compositae. Molec. Phy-
logenet. Evol.
xu — J. M. Porter, M. К. Wojciechows-
Domne 1995. The ITS
A valuable source of
Missouri Bot.
. S. Campbell & M. J.
region of nuclear ribosomal DNA:
evidence 'е оп — phylogeny. Ann.
Gard. 82: 247-277
P. | O. varieties and
Bally,
1959. Some new species,
Volume 89, Number 4
2002
Steinmann & Porter 481
Phylogenetic Relationships in Euphorbieae
forms of Monadenium (Euphorbiaceae). Candollea 17:
25- .
"1961. The Genus Monadenium. Benteli Pulish-
‚ Ber
Be ania. G. 1878. Notes on Euphorbiaceae. J. Linn.
Soc., Bot. 17: 185-267.
m Euphorbiaceae. Pp. 239-340 in G. Ben-
. Hooker (editors), Genera Plantarum 3.
Lovell — London
Berger, А. 1907 [1906]. Sukkulente Euphorbien. Eugen
Ulmer, —
Boissier, E. 1862. Euphorbieae. Pp. 3-188 in A. de €
Prodromus Systematis Naturalis, Pan
15(2). Vie d Masson & Fils, Paris.
— ——. 18 Peru ae (addenda et corrigenda). Pp.
р in А. de Candolle (editor), Prodromus Sys-
E matis Naturalis: Part 15(2). Victor Masson & Fils,
Paris.
Buck, W. & M. Hufi. 1977. Two new species of —
subge nus Agaloma from Mexico. J. Arnold Arbor.
343-348.
Bute һ, р 1971. d о |
509 in I. L. Wiggins . M. F Torá of
the Calg Is бш. Stanford Univ. Press, Stanford.
Carter, S. 1985. New species and taxonomic changes in
Euphorbia from east and northeast tropical Africa and
Oman. Kew A0: 809-825.
988a. Euphorbieae. Pp.
hill (editor), das of Tropic al * Africa, оонай:
aceae. А. /
-
587
4
er (editors).
a new с from
. Balkema, Rotterdam.
— ———,. 1988}. do qo section Somalica in Somalia.
PE ^ 6—38.
——— e species of Euphorbia subgenus Ti-
ruc alli ^ (Euphozbige eae) from Somalia and Oman. Nor-
dic J. Bot. 12: 675-679.
1992b. New herbaceous and woody species of
E oes ^ — eae) from Somalia. Nordic J. Bot.
12
— \ preliminary classification of е
subgenus Euphorbia. Ann. Missouri Bot. Gard. 81
366—319.
Croizat, E C. 1937. Notes of Euphorbiaceae, with a new
genus and a new subtribe of the Euphorbieae.
pine J. Sci. 64: 397 412.
———. |940. A significant new spec ies from New Guin-
ea: Euphorbia mmo Croiz., D nov. Bull. Jard.
Bot. Buitenzorg, ser. HI, 16: 351-35
1965. An моне НЫ ‘tion to ihe
sification of Euphorbia L., with
rican ar s M: ilagasy species. 1. Webbia 20: 573-706.
in introduction to the subgeneric clas-
sification s Euphorbia L; with stress on the South Af-
rican and pi agasy species. Ш. Webbia 27: 1-221.
Denis, M. 1921. Les Panhorst des ie
Tie. Pp. 66-69 їп
Philip-
—— тепе el аз-
Australes
Imprimerie Nemourienne,
rs
J. > 8 J. L. Doyle. 1987. A rapid DNA isolation
р > for ET — of fresh leaf tissue. Phy-
Su] Bull. -15
Dressler, R. L. ve ne res iE (Euphorbi-
aceae). Contr. Gray Heeb.
— — 62. A synopsis a тола (Е uphorbiaceae).
Ann. Missouri Bot. Gard. 48: 329—341.
Foster, P. I. 1994 Revision ү Euphorbia —
Teijsm. ex. Hassk. (Euphorbiaceae) and allies. Austro-
baileya 4: 227-24
Gilbert. M. G. 1987. co new geophytic species of Eu-
е — к оп the кк grouping of
African members. Kew Bull. 42: 231—244.
——— 199% . Six new species — A ege (subgenus
Esula) Dx "ED ges Kew Bull. 45: 265-276.
=== The relationships of ie Euphorbieae (Eu-
Missouri Bot. X
~
— eae). Ann.
Hassall, D. С. 1977. The genus RT in Australia.
Austral. г Bot. 25: 429-453.
Haworth, 1612. Synopsis Plantarum Succulentar-
um. Ric ha — Londor
House, H. 192
— of Me sit State.
Knfstated list of ferns and flowering
Bull. New York State Mus.
Huft E Ji 1497 9. A Monograph ч Euphorbia section Ti-
thymalopss Unpublished Ph.I |
of — igan, Ann Arbor.
84.
. Dissertation, Univer-
——— A review of Euphorbia (E — eae)
in Buc ш: Ann. Missouri Bot. Gard. 7 )21—
1027.
М. С. 1975.
of the Chihuahuan desert region and adjacent areas.
Wrightia : 120-1 "
Jordan, M. S. & W.
Lo testa in —— Collect.
21: 79-89.
Johnston, Studies of the Euphorbia species
1992. A survey of mu
Bot. (Barce dona) 2)
iden.
kim, K.-J. € R. K. Jansen. 1995. ndhF
lution and the major clades in А» e family.
Proc. Natl. Aci id. Sci., a : 10379-
Klotzsch, J. E. & nar 1860 Bos Hr.
Klotzsch las über Li inné's ens he Pflanzenklasse Tri-
rliner Herbarium's im Allge:
die natürliche Ordnung Euphorbiaceae insbesondere
а m Konig. Akad. Wiss. Berlin 1859:
36—
sequence evo-
coccae des Be nen und
Preuss.
Tricoc-
4 00. Linné's кү he Pflanzenklasse
ae. — . Wiss. Berlin 1859 [Phys. Ab-
hal
Koutnik. pm Chamaesyce (Eupl
newly onion genus in southern Addis 'a. S, Afric ап
J. Bot. 3: 262—
1987. : taxonomic revision of the Hawaiian
species of the genus Chamaesyce (Euphorbiaceae). Al-
331-388.
— 1:
C. d interesting taxa of the tribe
Africa
Leach, L. 1973. New an
Ы (Euphorbiac eae) from Portuguese
Garcia de Orta, Ser. Bot. 1: 31—42.
1957. Euphorbia mandravioky, nom. nov., el
un nom nouveau por une sous-section du genre Eu-
phorbe. Bull. Soc. Bot. France 104: 499—501.
Linnaeus, С. 1753. Species Plantarum. Impensis Lauren-
tii Salvii, dera m.
Mayfield, M. 1991.
aceae). a new
— johnstonii (E (Euphorbi-
aulipas. Mexico —
notes on Dahana н Асшае. Sida 11
9
species from Таш
57
1997. A Systematic Treatment of Euphorbia sub-
genus ‚Ренан (Е uphorbiaceae). Unpublished Ph.D
Disse — University of Texas, Austin.
Miller, P 17 The Gardeners Dictionary, Abridged 4th
ed. ne r At published), 1
Millspaugh, C. F. 1889.
Euphorbiaceae. Proc.
217-230.
1913. The
ew other = rican E uphorbiac eae.
ed 37
. Bot. 353-
London
C айин та to North Ame m
California Acad. Sci., 2nd ser.,
genera Pedilanthus = Cubanthus,
ield Mus. Nat
482
Annals of the
Missouri Botanical Garden
sh nt, D. 1996. Molecular Methods in Plant Biol-
y, 2nd ES Nic krent (self published), Carbondale.
— R. G. € J. A. Sweere. Combining data
in eileen systematics: An empirical approach us-
ing three vd cular data sets in Solanaceae. Syst. Biol.
43: 467-48
Park, K.-R. (7 6. Phylogeny of New World subtribe Eu-
ri (Euphorbiaceae). Korean J. Pl. Taxon. 26:
35-250.
1998. Monograph of Euphorbia n Tithymal-
opsis с eae). Edinbu — Bot. 55: 161-208.
—— & V. sens. 2000.
tribe Euphorbieae (Euphorbiaceae). Int. J. Pl. Sei. 161:
425434.
Pax, F 1894a. Euphorbiaceae africanae. Il. Bot. Jahrb.
19: 76-127.
1894b.
—— s A Mes Plantae Gürichianae.
Bot. Jahrb. 19: 142-143.
5:
. Euphorbiaceae. Рр. 1-168 in A. Engler
e (editors), Die Vegetation der Erde. IX. Die
n welt Afrikas. HI. 2. Wilhelm Engelmann, Leip-
Zg
. Hoffmann. 1931. Euphorbiaceae. Pp. 11-
233 11
ln sler & K. Prantl (editors), Die natürlichen
меана 2nd ed. 19C. Wilhelm Engelmann,
Leipzi
Pierre, ‘a 1896. Plantes du Gabon: 0 Bull.
Mens. Soc. Linn. Paris 1(159); 1259—
Planchon, P. Е, E а vrai nature Ps — des eu-
~
—
phorbes expliquée par nouveau genre
d'Euphorbiacées. Bull. Soc. Bot. j e 8: : -3:
Prokhanov, J. 1. 1933. Conspectus systematicus Tithy
malorum Asiae Medios. Trans. Rubber and a ha
nst. Moscow.
1949, Euphorbia. Pp. 304—495, 134—144 in Ko-
maros et al. (« editors), dos U.R.S.S
Rade Jiffe-Smith, A. The taxonomic ‘posited of Eu-
phorbia petiolata Sen p — n | of E. posti. (Eu-
phorbiaceae). Kew Bull. 29: 503-5
Vili study of
Rafinesque, C. S. 1837. Flora Telluriana 2. Rafinesque
self — as Iphia.
3. Flora Telluriana 4. Rafinesque (self pub-
lis hed), Pie lohia:
~
Rauh, W. ‚ Succulent and ae Plants of Mad-
agasc e p 2. Strawberry Press, Mill Valley.
Raven, P. H. 1963. мет ins ш relationships in the
floras s Py and South America. Quart. Rev. Biol.
38: 151-17
Steinmann, V. ү. 2001. The Evolution of Succulence in
the New World Species of Euphorbia (Euphorbiaceae).
Unpublished Ph. se e Claremont Graduate
Iniversity, Claremo
Swofford, D. L. 2000. PAUP, Phylogenetic Analysis Us-
ing Parsimony (*and Other Methods). Version 4. Sin-
and, Т husetts.
D. G. Higgins & T. J. Gibson. 1994.
ISTAL W: — ving the sensitivity of progressive
AU i sequence alignment through sequence weight-
gap penalties and weight matrix
Acids Res. 22: 4673-4680.
Trew, €. $ 1754. Herbarium Blackwellianum, Vol.
pen у,
Webster, С. | 967. The genera of Euphorbiaceae in the
southeastern United States. J. Arnold Arbor. 48: 303—
361
—
—
5
=
2°
—s
=
جس
-
У.
en
o
=
=
—
p — c onspectus o a new classification of the
Е eae. Taxon 24: 593—601
——. 1994. Synopsis of the genera and раа
taxa of l Euphorbia eae. Ann. Missouri Bot. Gard.
33-14
t wn & B. N. Smith. 1975. Systematics
of synt — fixation pathways in Euphor-
bia. Tuo n 24: 27-33.
. Rupert & D. Koutnik. 1982. Systematic sig-
niña ‘ance of pollen nuclear number in En ласеае,
tribe epee Amer. J. Bot. 69: 407—
Wheeler, 2 1943. The genera of ne —
X50
Amer. B de 30: 45€
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
ppe
"i cultivated * Get ed taxa, the native origin of the plant is given in parentheses
ndix |.
icher information for the included species of Euphorbieae and ойон sequenced in this study.
ITS ndhk
GenBank GenBank
Taxon Origin and voucher accession accession
OUTGROUPS
Omalanthus populifolius Graham Cultivated (nat. Australia), Steinmann AF537585 AF538202
23 (RSA)
Sapium sebiferum (L.) Roxb. Cultivated (nat. China). Steinmann 1424 AF537580 AF538261
Sebastiania cornuta McVaugh Mexico, Sonora, Steinmann 589 (RSA) AF537587 AF538263
Stillingia spinulosa Torr. Mexico, Sonora, Felger 92-381 (RSA) AF537588 AF538264
EUPHORBIEAE SUBTRIBE ANTHOSTEMINAE
Anthostema madagascariense Baill. Madagascar, Pascal 586 (MO) AF537582 AF538257
Anthostema senegalense A. Juss. Senegal. Bamps 7759 (MO) — AF538259
Anthostema sp. nov. — ar, Miller et al. 8840 (MO) AF537583 AF538258
Dichostemma glaucescens L. Pierre Gabon, McPherson 15531 (DAV) AF537584 AF538200
EUPHORBIEAE SUBTRIBE NEOGUILLAUMINIINAE
Calycopeplus casuarinoides L.S. Sm. Cultivated (nat. Australia). Steinmann AF537580 -—
7 (RSA)
Calycopeplus collinus РЛ. Foster Australia, гап der Werff 11848 (DAV) — AF538254
Calycopeplus paucifolius (Klotzsch) Australia, Craven 7139 (RS — AF538255
Baill.
Neoguillauminia cleopatra (Baill.) Cro- New Caledonia. WePherson 17882 (MO) AF537581 AF538250
izat
EUPHORBIEAE SUBTRIBE EUPHORBIINAE
Chamaesyce acuta (Engelm.) Millsp. U.S.A.. Texas. Mayfield 1989 (RSA) AF537450 AF538176
Chamaesyce angusta (Engelm.) Small
Chamaesyce articulata (Burm.) Britton
Chamaesyce carunculata (Waterf.) Shin-
ners
Chamaesyce degeneri (Sherff) Croizat &
Degener
Chamaesyce hypericifolia (L.) Millsp.
Small
Chamaesyce polycnemoides (Boiss.) J.
м
Chamaesyce prostrata (Aiton
Soják
Endadenium gossweileri (N.E. Br.) L.C.
Leach
— aaron-rossú A.H. Holmgren
¿ N.H. Holmgren
Euphorbia abdelkuri Balf.f.
Euphorbia acalyphoides Hochst. ex
0188.
Euphorbia acanthothamnus Heldr. &
Sart. ex Boiss.
Euphorbia adiantoides Lam.
Euphorbia agowensis Hochst. ex Boiss.
Euphorbia alluaudii Drake
Euphorbia alta Norton
Euphorbia amygdaloides L.
Mexico, акак
U.S.A.,
i
Mayfield 1328 (RSA)
irgin Isalnds. Steinmann 94-10
)
Mexico, Chihuahua. Bowers et al. 2939
(ARIZ)
U.S.A.,
U.S.A.,
U.S.A..
Hawaii.
Hawaii.
Hawaii.
Motley 338 (NY)
Motley 1804 (NY)
Motley 1802 (NY)
Botswana. Snow «€ Chatakuta 6905
(MO)
Cultivated (nat.
SA)
U.S.A.. Arizona. Ross s.n. (RSA)
Cultivated (nat.
SA)
Kenya. Luke et al. TPR177 (MO)
Angola). Steinmann
Abd-el-Kuri). Steinmann
Cultivated — Greece/Turke у), Stein-
mann 142.
)
Ecuador. pita & Anderson 22548
(G
Ethiopia, Gilbert & Thulin 132 (MO)
Cultivated (nat. Madagascar). Steinmann
1481 (RSA
U.S.A.
)
. Arizona. Sanders 5905 (RSA)
C — d * Europe). Steinmann
1428 (RS.
AF537449
AF537446
AF537447
AF537444
AF537443
AF537445
AF537448
AF537471
AF537396
AF537458
AF537395
AF537419
AF537468
AF537553
AF538175
AF538198
AF538194
AF538251
AF538232
AF538197
Annals of the
Missouri Botanical Garden
Appendix |. Continued.
ITS ndhk
GenBank GenBank
Taxon Origin and voucher accession accession
Euphorbia ankarensis Boiteau Cultivated (nat. Madagascar), Steinmann AF537462 —
2 (RSA)
Euphorbia antisyphilitica Zucc. Cultivated (nat. U.S.A., Texas), Stein- AF537398 —
mann 1451 (RSA)
Euphorbia antso Denis Cultivated — ии ar), Steinmann AF537579 AF538253
1473-B (RS
Euphorbia aphylla Brouss. ex Willd. Cultivated va. A anary Islands), Stein- AF537540 AF538225
mann 1466 (RSA)
Euphorbia appariciana Колт Cultivated ina Brazil), Steinmann 1442 AF537455 AF538177
(RSA)
Euphorbia arbuscula Balf.f. Cultivated (nat. Socotra), Steinmann AF537496 —
1435 (RSA)
Euphorbia aff. ariensis HBK Mexico, Nayarit, Steinmann 1148 (RSA) AF537409 —
Euphorbia atrispina N.E. Br. Cultivated (nat. Republic of S. Africa), AF537568 —
Steinmann 1478 (RSA)
Euphorbia atropurpurea Brouss. ex Cultivated (nat. Canary Islands), Stein- AF537542 AF538230
Willd. mann 1489 (RSA)
Euphorbia attastoma Rizzini Cultivated (nat. Brazil), Steinmann 1487 AF537511 —
(RSA)
Euphorbia balsamifera Aiton ssp. ade- Cultivated (northeast tropical Africa), AF537571 AF538250
nensis (Deflers) Bally а 1480 (RSA)
Euphorbia bicolor Engelm. & A. Gray . Texas, Van Devender 96-290 AF537380 —
pes
Euphorbia bifurcata Engelm. Mexico, Nuevo León, Nesom 7703 AF537434 AF538173
SA)
Euphorbia bilobata Engelm. U.S.A.. Arizona, Steinmann 938 (RSA) AF537435 AF538172
Euphorbia boöphthona C.A. Gardner Australia, Coveny 3054 (RS AF537515 AF538207
Euphorbia brunellii Chiov. Cultivated (nat. east аре 'al Africa), AF537480 AF538203
Steinmann 1495 (RSA)
Euphorbia calcicola Fern. Mexico, Morelos, Steinmann 801 (RSA) AF537385 —
Euphorbia californica Benth. Cultivated (nat. Mexico), Steinmann AF537377 —
92 (RSA)
Euphorbia calyculata HBK Mexico, Puebla, d L. 7261c (NY) AF537524 AF538221
Euphorbia calyptrata Coss. € Kralik Morocco, Podlech - 8 (RSA) AF537549
Euphorbia capmanambatoensis Rauh Cultivated (nat. oe ar), Steinmann AF537470 —
1468 (RSA)
Euphorbia caputmedusae L. Cultivated (nat. Republic of S. Africa), AF537574 —
Steinmann 1463 (RSA)
Euphorbia cassythoides Boiss. Cayman Islands, Proctor 47858 (NY) AF537387 —
Euphorbia ceroderma He Johnst. Mexico, Sonora, Steinmann 1393 (RSA) AF537389 AF538153
Euphorbia cestrifolia | Ecuador, Harling 27200 (GB) AF537521 AF538213
пако cheirolepis * h. & С.А. Central Asia, Vasák s.n. Ms AF537424 —
Mey
ооб chersonesa Hult Mexico, Baja California Sur, Steinmann АЕ537436 AF538174
1252 (RS
Euphorbia clava Jacq. Cultivated (nat. Republic of S. Africa). AF537509 AF538245
Steinmann 1472-B (RSA)
Euphorbia colletiodes Benth. Mexico, Sonora, Steinmann 93-387 AF537405 —
(ARIZ)
Euphorbia comosa Vell. Brazil, Webster 25425 (DAV) AF537503 AF538222
Euphorbia crossadenia Pax & K. Hoffm. Brazil, Graças et ds 886 (SP) AF537451 —
Euphorbia crotonoides Boiss. Tanzania, M.R. 23765 (NY) AF537578 AF538238
Mexico, — Steinmann 1199 AF537393 AF538152
Euphorbia delicatula Boiss.
(RSA)
Volume 89, Number 4
2002
Steinmann & Port
485
Phylogenetic э Н in Euphorbieae
Appendix |. Continued.
ҺЕ
GenBank GenBank
‘Taxon Origin and voucher accession accession
Greece, Strid 25582 (RSA) AF537539 —
Euphorbia dendroides L.
Euphorbia denisii Oudejans
Euphorbia depauperata Hochst. ex A.
Rich.
Euphorbia discolor Ledeb.
Euphorbia drupifera Thonn.
Euphorbia eanophylla Croizat
Euphorbia eglandulosa V.W. Steinm.
Euphorbia elata Brandegee
Euphorbia elliotii Leandri
Euphorbia epiphylloides Kurz
Euphorbia equisetiformis A. Stewart
Euphorbia eremophila A. Cunn.
Euphorbia eriantha Benth.
Euphorbia esculenta Marloth
Euphorbia espinosa Pax
Euphorbia esula L.
Euphorbia exstipulata Engelm.
Euphorbia fulgens Karw. ex Klotzsch
Euphorbia gentryi V.W. Steinm. & T.F.
Daniel
Euphorbia germainii Phil.
Euphorbia geroldii Rauh
Euphorbia glanduligera Pax
Euphorbia globosa (Haw.) Sims
Euphorbia goetzei Pax
Euphorbia gollmeriana Klotzsch ex
Boiss.
Euphorbia gottlebei Rauh
Euphorbia gradyi V.W. Steinm. & A.
Ram.-Roa
Euphorbia graminea Jacq.
Euphorbia grantii Oliv.
Euphorbia gregaria Marloth
Euphorbia guatemalensis Standl. &
Steyerm.
E uphorbia guerichiana Pax
Euphorbia gymnoclada Boiss.
Euphorbia gymnonota Urb.
Cultivated (nat. Madagascar), Steinmann
1434 (RSA)
Malawi. Kaunda & Tawakali 906 (NY)
Former U.S.S.R., McNeal 710 (RSA)
Cultivated (nat. Africa), Steinmann 1488
(RSA)
Bolivia, Beck 11399 (DAV)
хїсо, ‚© hiapas, Breedlove 70137
(С
ы m 6694 (NY)
Madagascar, Dorr et al. 3985 (MO)
Cultivated (nat. Andaman Islands),
Steinmann 1459 (RSA)
Ecuador, Galapagos Islands, Eliason &
Eliason 1573 (К)
Australia, Vasek 680914-51 (RSA)
U. Arizona. Steinmann 925 (RSA)
Cultivated (nat. Re — of S. Africa).
Steinmann 1474 (RSA)
Cultivated (nat. ne a), Steinmann 1494
(RSA)
U.S.A.. New Jersey (nat. Eurasia), Stein-
mann 1427 (RSA)
U.S.A.. Arizona, Steinmann 934 (RSA)
Mexico, Oaxaca. Campos 813 (RSA)
Mexico, Sonora, Steinmann 94-357
(ARIZ)
Chile. Teillier 4267 (SGO)
Cultivated (nat. Madagascar). Steinmann
1467 (RSA)
Namibia, Koutnik 2015 (DAV)
Cultivated (nat. Re pacc of S. Africa).
Steinmann 1454-A (RS
Tanzania, Taylor et if 6490 (MO)
Venezuela, Ramírez 2696 (DAV)
Cultivated (nat. Madagascar), Steinmann
71 (RSA)
Mexico, Oaxaca, Steinmann 784 (RSA)
Mexico, Sonora, Steinmann 94-107
(RSA)
Tanzania, Bidgood et al. 1186 (MO)
Cultivated (nat. Republic of S. Africa),
Steinmann 1445-B (RSA)
Mexi hiapas, Steinmann 1170
HE.
— of South Africa, Balkwill et al.
6022 (MO)
Brazil, Webster 25853 (DAV)
Bahama Islands, Correll & Wasshausen
46769 (NY)
AF537497
AF537547
AF537480
AF537498
AF537394
AF537495
AF537478
AF537484
AF537388
AF537423
AF537440
AF537416
AF537540
AF537433
AF537404.
AF537400
AF537499
AF537475
AF537420
AF537413
AF537501
AF537459
AF537407
AF537410
AF537527
AF537408
AF537415
AF537450
AF537507
AF538191
AF538217
AF538107
AF538190
AF538229
AF538171
AF538154
AF538205
AF538178
AF538239
AF538185
AF538220
AF538151
AF538155
AF538242
AF538182
Annals
of the
Missouri Botanical Garden
Appendix l. Continued.
ITS idhk
GenBank GenBank
Taxon Origin and voucher accession accession
Euphorbia haeleeleana К R. Herbst U.S.A., Hawaii, Fernstemacher s.n. (NY) AF537514 AF538206
Euphorbia hallii R.A. Cultivated (nat. Re public of S. Africa), AF537573
Steinmann 1475 (RS!
Euphorbia hamata (Haw.) Sweet Cultivated (nat. Republic of S. Africa), — AF538237
Steinmann 1454-B (RSA)
Euphorbia hedyotoides N.E. Br. Cultivated (nat. Madagascar), Steinmann AF537460 AF538196
2-A (RSA)
Euphorbia heterodoxa Müll. Arg. Brazil, Webster 25810 (DAV) AF537500 —
Euphorbia heterophylla L. Mexico, Sonora, Van Devender 98-1157 AF537429 AF538170
(ARIZ)
Euphorbia — (Klotzsch & Costa Rica, Haber 10501 (F) AF537508 AF538211
Garcke) B
Euphorbia — Radel.-Sm. Mexico, Tamaulipas, Mayfield & Patter- AF537431 AF538165
son 1843 (TE)
Euphorbia iharanae Rauh Cultivated (nat. Madagascar), Steinmann AF537477 —
1458 —
Euphorbia innocua LC. Wheeler U.S.A., Texas, — ld 2168 (RSA) AF537380 AF538161
Euphorbia insulana Vell. Brazil, ns 03 (N AF537411
Euphorbia ipecacuanhae L.. U.S.A., South ( cali Spongberg & AF537397 —
Boufford 1718 (MO)
Euphorbia jaliscensis B.L. Rob. & Mexico, Jalisco, Steinmann 754 (RSA) AF537442 AF538160
Greenm.
Euphorbia juttae Dinter Cultivated (nat. Namibia), Steinmann AF537418 AF538188
Euphorbia kraussiana Bernh. و of S. Africa, Stalmans 372 AF537548 —
О)
Euphorbia lacera Boiss. Mexico. К. de Mexico, Castilla & Tejero AF537441 —
958 (ENCB
Euphorbia lactiflua Phil. Chile, Dillon & Teillier 5105 (F) AF537528 AF538219
Euphorbia lagunensis Hufi Mexico, Baja California Sur, Steinmann AF537379 —
272 (RSA)
Euphorbia lagunillarum Croiz. Cultivated (Venezuela), Steinmann 1621 AF537502 —
(RSA)
Euphorbia lathyris L. U.S.A., California (nat. Eurasia), Stein- AF537550 —
mann 1426 (RSA)
Euphorbia laurifolia Juss. Ecuador, Mena C61 (NY) AF537509 —
Euphorbia leucocephala Lotsy Cultive ate d (nat. Guatemala), Steinmann AF537381 —
94-17 (RSA
Euphorbia lignosa Marloth C abes d (nat. Namibia), Steinmann — AF538240
455 (RSA)
Euphorbia longifolia Lam. Canary Islands, Legaard 9905 (DAV) AF537558 AF5382:
Euphorbia longituberculosa Boiss. Cultivated (nat. east tropical Africa), AF537577 АЕ ien
Steinmann 1479 (RS!
Euphorbia macropus (Klotzsch & Gar- Mexico, Nuevo León, Mayfield 1294 AF537378 —
‘ke 0188. ( T EX)
Euphorbia macvaughii Carvajal & Lo- Cultivated (nat. Mexico, Michoacán), AF537382 —
melí Steinmann 1486 (RSA)
Euphorbia mahabobokensis Rauh Cultivated (nat. Madagascar), Steinmann AF537522 —
1456 (RSA)
Euphorbia matabelensis Pax Botswana, Smith 4229 (MO) — AF538247
Euphorbia mauritanica L.. Cultivated (nat. southern Africa), Stein- AF537531
mann 1432 (RSA)
Euphorbia medicaginea Boiss. Morocco, Podlech 41883 (RSA) AF537535 —
Euphorbia meenae S. Carte India, Singh s.n. (RSA) AF537483 AF538202
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
Appendix l. Continued.
ITS ndhl
GenBank GenBank
Taxon Origin and voucher accession accession
Euphorbia megalatlantica Ball Morocco, Podlech 41177 (RSA) АЕ537536 AF538226
Euphorbia meloformis Aiton Cultivated (nat. Republic of S. Africa), AF537505
Steinmann 1490 (RSA)
Euphorbia meuleniana O. Schwartz Cultivated (nat. Yemen), Steinmann AF537572 —
1448 (RSA)
Euphorbia milii Des Moul. Cultivated (nat. Madagascar), Steinmann AF537461 AF538195
1476 (RSA)
Euphorbia millotii Ursch & Leandri Cultivated _ Madagascar), Steinmann AF537403 —
1477 (R
Euphorbia misella S. Watson Mexico, C — Steinmann 1032 AF537384 AF538160
(RSA)
Euphorbia misera Benth. Mexico, Baja California, Steinmann AF537383
1285 (RSA)
Euphorbia monteiri Hook.f. RS Long & Rae 290 (K) AF537503 —
Euphorbia myrsinites L. Colorado (nat. Eurasia), Stein- AF537551 AF538231
mann 1430 (RSA)
Euphorbia namuskluftensis L.C. Leach Cultivated (nat. Bebo: Steinmann AF537502 —
191 (RSA)
Euphorbia oaxacana B.L. Rob. & Mexico, Colima, Steinmann 1094 (RSA) AF537373 —
Greenm.
Euphorbia obesa Hook.f. Cultivated (nat. Re public of S. Africa), AF537566 AF538244
ips inmann 1465 (
Euphorbia oblongata Griseb. U.S.A., California (nat. Europe), Halse AF537555
4. : 34 (К;
Euphorbia ocymoidea L. Mexico, Jalisco, Steinmann 1139 (RSA) AF537392
Euphorbia oerstediana (Klotzsch € Gar- U.S.A., Puerto Rico, Axelrod & Sastre AF538159
0185. 6278 (NY)
Euphorbia omariana M.G. Gilbert Ethiopia, Friss et al. 3674 (К) AF537500 AF538243
nd panchganiensis Blatt. & India, Singh s.n. (RSA) AF537375 —
Са
Euphorbia papillosa A. St.- Argentina, Renvoize 3068 (RSA) AF537510 —
Euphorbia pedilanthoides a Cultivated (nat. Madagascar), Steinmann — AF538192
38 (RSA)
Euphorbia pentadactyla Griseb. Argentina, Cantino 700 (. han, AF537428 —
Euphorbia peperomioides Boiss. Brazil, Nakjima et al. 1596 (DAV) AF537523 —
Euphorbia peplus U.S.A., California (nat. E. Stein- AF537532 AF538228
mann 1433 (RS
Euphorbia perrieri Drake us d (nat. Madagascar), Steinmann AF537463 =
3 (RSA)
Euphorbia pervilleana Baill. — (nat. Madagascar), Steinmann AF537518 AF538209
1444 (RSA)
Euphorbia petiolata Banks & Sol. Middle East, Liston 7-85-388/3 (RSA) AF537422 AF538180
Euphorbia phosphorea Mart. AF537512 AF538223
Euphorbia phylloclada Boiss.
Euphorbia pilosa L.
Euphorbia pirottae N. Terrac.
Euphorbia platycephala Pax
Euphorbia platyclada Rauh
Euphorbia plumerioides Teijsm. ex
Hassk
Cultivated (nat. Brazil), Steinmann 1446
RSA)
Republic of S. Africa, Oliver et al. 6611
K
Former U.S.S.R., Elias et al. 7182
(RSA)
Cultivated (nat. '
1440 (RSA)
Tanzania, Bidgood et al. 2249 (К)
Cultivated (nat. Madagascar), Steinmann
147 (RSA)
Australia, Fryxell & Craven 4022 (RSA)
Tanzania), Steinmann
AF537427
AF537557
AF537417
AF537501
AF537421
AF537513
AF538179
AF538234
AF538180
AF538241
AF538187
488
Annals of the
Missouri Botanical Garden
Appendix l. Continued.
GenBank GenBank
Taxon Origin and voucher accession accession
Euphorbia poissonii Pax Cultivated (west Piu ral Africa), Stein- AF537482 AF538193
mann 1498
Euphorbia polyantha Pax Kenya, Faden & — 74/477 (К) — AF538189
Euphorbia primulifolia Baker Cultivated (nat. Madagascar), Steinmann AF537466
1 (RSA
Euphorbia pteroneura A. Berger (1) Mexico, Chiapas, Mayfield 980 (TEX) AF537506 AF538210
Euphorbia pteroneura A. Berger (2) Cultivated (nat. Mexico/Cent. America), AF537505
Steinmann 1622 (RSA)
Euphorbia pteroneura A. Berger (3) Cultivated (nat. Mexico/Cent. America), AF537504 —
Steinmann 1623 (RS.
Euphorbia pulcherrima Willd. ex Mexico, Nayarit, Steinmann 1070 (RSA) AF537432 AF538168
Klotzsch
Euphorbia pumicola Hufi ee xico, Baja California Sur, Steinmann AF537437 AF538164
1 (ARIZ)
Euphorbia punicea Sw C — d (nat. Jamaica), Raz 193 (NY) AF537516 AF538208
Euphorbia radians Be nth. exico, Sonora, Steinmann 944 (RSA) AF537438 AF538169
Euphorbia regis-jubae Webb & Berthel. C шый (nat. Canary Islands), Stein- AF537541
mann 1431 (RSA)
Euphorbia rhombifolia Boiss. Cultivated (nat. Re "public of S. Africa), AF537414 AF538183
Steinmann 1439 (RSA)
Euphorbia robusta (Engelm.) Small .5.А., Colorado, — 1429 AF537533 AF538227
(RSA)
Euphorbia rossiana Pax Mexico, Puebla, Steinmann 1195 (RSA) AF537374 —
Euphorbia rossiana var. nov. Mexico, Guerrero, Steinmann 1199 AF537390 —
(RSA)
Euphorbia rossii Rauh & Buchloh Cultivated — Madagascar), Steinmann AF537465 —
1449 (RS
Euphorbia rubella Pax Cultivated (ninh east tropical Africa), AF537487 AF538204
Steinmann 1464 (RSA
Euphorbia rzedowskii Mc Vaugh Mexico, Michoacán, Steinmann 1110 AF537399 —
(RSA)
Euphorbia sarcodes Boiss. Brazil, Cordeiro et al. 2233 (SP) AF537454 —
Euphorbia scatorhiza S. Carter Cultivated (nat. Somalia), Steinmann AF537420 AF538181
1441 (RSA)
Euphorbia scheffleri Рах Cultivated (nat. Kenya), Steinmann 1452 — AF538249
RSA)
Euphorbia schimperi Presl Cultivated (nat. Arabian Peninsula/NE 11537537 —
tropical Africa), Steinmann 1499
(RS/
Euphorbia sessilifolia Klotzsch ex Boiss. Brazil, Arbo 5418 (DAV) AE537453 —
)
Euphorbia sessilifolia Klotzsch ex. Boiss. Brazil, Cordeiro et al. 2218 (SP) AF537452 —
2
Euphorbia segoviensis (Klotzsch & Gar- Mexico, Chiapas, Steinmann 1174 AF537400 —
cke) Boiss. RS/
Euphorbia sinaloensis Brandegee Mexico, Sonora, Steinmann 94-130 AF537401 AF538156
Euphorbia sipolisii N.E. Br Cultivated (nat. Brazil), Steinmann 1443 AF537517 —
(RSA)
Euphorbia socotrana Balf.f. Cultivated A Socotra), Steinmann AF538248
1436 (RS
Euphorbia sonorae Rose Mexico, id Steinmann 964. (RSA) AF537402 —
Euphorbia soongarica Boiss. Former U.S.S.R., Elias RSA) AF537545 —
Euphorbia spathulata Lam. U.S.A., California, Banks 1601 (RSA) AF537552 AF538233
Euphorbia sphaerorhiza Benth. Mexico, Sonora, Steinmann 1020 (RSA) AF537412 AF538158
Volume 89, Number 4
2002
Steinmann & Porter
Phylogenetic Relationships in Euphorbieae
Appendix l. Continued.
ITS
GenBank GenBank
Taxon Origin and voucher accession accession
AF537529 —
Euphorbia stenophylla (Klotzsch & Gar-
cke) Boiss.
Euphorbia stricta L.
Euphorbia strigosa Hook. & Arn.
Euphorbia subpeltata 5. Watson
Euphorbia succedanea L.C. Wheeler
Euphorbia tannensis Spreng.
Euphorbia tanquahuete Sessé & Mociño
Euphorbia teke Schweinf. ex Pax
Euphorbia tetraptera Baker
Euphorbia thinophila Phil.
Euphorbia thouarsiana Baill.
Euphorbia tirucalli L.
Euphorbia trichadenia Pax
Euphorbia trichotoma HBK
Euphorbia tuberosa L.
Euphorbia tubiglans Marloth ex R.A.
Euphorbia turczaninowii Kar. & Kir.
Euphorbia usambarica Pax
Euphorbia weberbaueri Mansf.
Euphorbia whitei L.C. Wheeler
Euphorbia zonosperma Múll.A
Euphorbia xylophylloides Brongn. ех
Lem.
Euphorbia sp.
Euphorbia sp. nov. |
Euphorbia sp. nov. 2
Monadenium elegans S. Carter
Monadenium ellenbeckti N.F. Br.
Monadenium lindenii S. Carter
Monadenium magnificum E.A. Bruce
Pedilanthus bracteatus Jacq.
Pedilanthus calcaratus Schltdl.
Pedilanthus connatus Dressler & Saca-
mano
Pedilanthus cymbiferus Schltdl.
Brazil, Aparecida da Silva & dos Santos
3267 (DAV
Austria, Wallnofer 8531 (NY)
Mexico, Nayarit, Steinmann 1079 (RSA)
Mexico, Morelos, Steinmann 794 (RSA)
Mexico, Aguascalientes, Steinmann 745
(RSA)
Australia, Fryxell et al. 4475 (RSA)
Cultivated (Mexico), Steinmann 1620
RSA
Cultivated (nat. east tropical Africa),
Steinmann 1470 (RSA
— ar, Rapes ohita — (DAV)
Chile, Dillon & Teillie 7 (Е
— аг, pda & Ranaivoja-
ona 14565 (K)
Cultivated (nat. Africa/Madagascar),
Steinmann 1445-A (RSA)
Cultivated (nat. Zimbabwe/Angola),
Steinmann 1461 (RS!
Belize. Hill 20357 (MO)
C — d (nat. dye of S. Africa),
Steinmann 1472 USA)
Cultivated (nat. Re publi of S. Africa),
Steinmann 1462 (RS
China, Liston 827-4 (RS, А)
Tanzania, Balslev 291 (NY
Cultivated (nat. Peru), Steinmann 1347
(RSA
Mexico, Oaxaca, Torres 10833 (DAV)
Brazil. Plowman et al. 8579 (F)
Cultivated (nat. Madagascar), Steinmann
1450 (RSA)
Cultivated (nat. probably Africa), Stein-
mann 1469 (RSA)
Brazil. Coredeiro et al. 2203 (SP)
Tamaulipas, Mayfield 1851
| lexico.
Cultivated (nat. Tanzania), Steinmann
1473-A (RSA
Cultivated (nat. east ida “al Africa).
Steinmann 1453 (RS
Cultivated (nat. Somalia), Steinmann
1485 (RSA)
Cultivated
1496 (RS
— d (nat. Mexico). Steinmann
1460 (RSA)
— о, a raeruz, Cházaro B. & de
Cházaro 7294 (NY)
Mexico. Jalisco, Sacamano s.n. (MO)
Tanzania), Steinmann
Mexico, Puebla, Steinmann 1624 (RSA)
AF537559
AF537439
AF537376
AF537403
AF537425
AF537525
AF537485
AF537526
AF537530
AF537474
AF537504
AF537507
AF537543
AF537391
AF537430
AF537407
AF537481
AF537457
AF537470
AF537472
AF537489
AF537492
AF537493
AF537491
AF538163
AF538162
AF538184
AF538224
AF538218
AF538230
AF538246
AF538212
AF538214
AF538199
AF538200
490
Annals of the
Missouri Botanical Garden
Appendix l. Continued.
ITS idhF
GenBank GenBank
Taxon Origin and voucher accession accession
Pedilanthus finkii Boiss. Mexico, Oaxaca, Meave del Castillo AF537520 —
1551 (MO)
Pedilanthus macrocarpus Benth. Mexico, Baja California, Steinmann AF537490 —
235 (RS
Pedilanthus tehuacanus Brandegee Mexico, Puebla, Steinmann 1400 (RSA) AF537488 AF538215
Pedilanthus tithymaloides (L.) Poit. Guatemala, Castillo 2713 (NY) AF537494 АЕ538216
Synadenium grantii Hook.f. Cultivated (east tropical Africa), Stein- AF537469 AF538201
mann 1497 (
—
~~
SA)
MOLECULAR Jason C. Bradford?
PHYLOGENETICS AND
MORPHOLOGICAL
EVOLUTION IN CUNONIEAE
(CUNONIACEA E)!
ABSTRACT
The Cunonieae are the largest tribe in the flowering plant family Cunoniaceae and include the widespread genus
Weinmannia. This study aims to understand phylogene lic re lationships within Cunonieae e by using DNA sequences in
a parsimony-cladistic analysis. Sequence ph loci included the internal transcribed s (ITS-1 ind ITS- 2) of nuclear
е дерине | DNA. and the tral. intron and trnl-F spacer of chloroplast DNA. жайы p» aaa ific amplification of
non-orthologous ITS-2 copies made it necessary to exclude the IT 5-2 data, but otherwise the nuclear and ¢ : 0
ala sels were congruent. The results place Vesselowskya as the sister genus to other Cunonieae and support the
нат of Panc 5 Cunonia, and all five sections of Weinmannia, but do not ae how these а аге
related. The os of Weinmannia sect. Weinmannia is upheld, with W. trichosperma from temperate forests of
South America and W. tinctoria from the Mascarene Islands placed basal to a large clade of tropical American species.
Although e maintain the monophyly of Weinmannia, this is neither verified nor statistically refuted
by the molecular data. Likewise, Cunonia, with one isolated South African species, has only weak molecular support
but ¢ = morphological ee | Lack of support for relationships among major clades within Cunonieae makes
it difficult to suggest patterns of ораев, ‘al evolution. However. a well-qupportid phylogenetic hypothesis within
инна 5 sect. Leiospermum is used to discuss hete тоору in inflorescence architecture. Uniquely derived features
of the inflorescence are found in the New Caledonian species Weinmannia dichotoma and in the New Zealand species
W. silvicola and W. racemosa. These heterotopic changes involve alternate patterns in the fate of terminal meristems
and s arrangement of metamers bearing racemes. In an appendix the correct orthographies and original publications
of all five sections of Weinmannia are жейде: types are also designated for Weinmannia sections /nspersae. and
сае in order to — al em.
Key words: cladistics, Cunonia, Cunoniaceae, Cunonieae, evolution, Fasciculatae, heterotopy. гон Ак e archi-
lecture, — sae, ITS. J a molecular systematics, Pancheria, paralogous loci, Spicatae, trnk. ‚ Vesse-
lowskya, Weinmannia.
The flowering plant family Cunoniaceae R. Br. 1998a). Weinmannia is divided into five sections,
(Oxalidales) (Angiosperm Phylogeny Group, 1998) with each one largely restricted to a particular geo-
comprises about 300 species in 26 genera. Plants graphic region. The Cunonieae are also composed
f the two other largest genera in Cunoniaceae,
of the family are trees and shrubs in wet tropical
and cool temperate forests, with most genera oc- Pancheria, with about 30 species endemic to New
curring in eastern Australia, Melanesia, and New Caledonia (Guillaumin, 1940; Morat, 1993), and
Cunonia, with about 25 species in New Caledonia
Guinea. About 210 Cunoniaceae species are in ¿
monophyletic group of four genera called the tribe and 1 species in the South African Cape region
Cunonieae (Bradford & Barnes, 2001). Weinmannia (Hoogland et al.. 1997). Vesselowskya. the remain-
is by far the largest and most widely distributed ing genus in the tribe, has only two species endem-
genus, with over 150 species found in the Ameri- ic to eastern Australia (Rozefelds et al., 2001).
cas, islands of the eastern Indian Ocean, Malesia. A few recent publications have provided new in-
and the South Pacific (Bradford, 1998: Hopkins. sights on the taxonomy and phylogeny of Cunon-
' This research has been supported by a U.S. National Science Foundation — grani (BIR-9250779) to Wash-
ington University; a Mellon Foundation * к the Missouri Botanical Garden; a U.S. National Science Foundation
Dissertation Improvement Grant (#5747 Э) to В. Schaal and J. Bradford: and а travel grant from the American Society
of Plant Taxonomists to J. Bradford. " с A1 or the ueri and ms of many colleagues from institutions
ا the world, including: AKU, ОБУР, COL. FRIM, LOJA, MA OU, NSW, P, PORT, QCNE, SAN, SAR,
5 , TAN, TEF, UNIMAS, USZ, V И i. and the ^ etii p iety (W. E to B. Schaal for work in her laboratory
al edes University; and to C. Feuillet, V. Hollowell, H. Fortune Hopkins, J. Rauscher, and A. Rozefelds for help
on the manuscript.
? Research Scientist, Missouri Botanic z Garden. Visiting Sc ‘holar, мү Science and Policy, 2132 Wickson
Hall. University of California at Davis, Davis, California 95616, U.S.A. bradford@ice.ucdavis.edu.
ANN. MISSOURI Bor. GARD. 89: 491-503. 2002.
492
Annals of the
Missouri Botanical Garden
ieae. In a study of the relationship between Wein-
mannia and its putative close relative, Cunonia,
using a cladistic analysis of morphological features
(Bradford,
monophyly of Weinmannia, four of its sections, and
1998) there was weak support for the
of Cunonia. A family-level analysis by Bradford
and Barnes (2001) based on morphology and chlo-
roplast DNA sequences (rbcL and trnL-trnF) estab-
lished a new tribal classification, and clarified ge-
neric. circumscriptions by proposing apomorphic
morphological characters for each genus.
Several recent generic revisions have provided
valuable details on taxonomic distribution and mor-
phology for many Cunonieae species. Rozefelds et
al. (2001) gave a table summarizing similarities and
differences among genera of Cunonieae and de-
scribed a new species of Vesselowskya. New species
descriptions and keys have also been produced for
Cunonia (Hoogland et al., 1997) and Malagasy
Weinmannia (Bradford, 2001; Bradford & Miller,
)01). Revisions have been completed for Wein-
mannia of Malesia and the South Pacific (Hopkins,
1998a, b, c; Hopkins & Florence, 1998). Weinman-
nia of the Americas are poorly studied in their en-
tirety, although some national and regional treat-
ments have been done (Harling, 1999; Bradford &
Berry, 1998).
Despite these efforts, relationships among Cu-
nonieae genera are unclear, the monophyly of Wein-
mannia is in doubt, and the monophyly of some
sections within Weinmannia is poorly established.
In this study, I use DNA sequences from chloro-
plast and nuclear loci to clarify phylogenetic rela-
tionships within Cunonieae. In addition, I show that
the phylogenetic hypotheses generated by this mo-
lecular data can help reevaluate character evolu-
tion within the tribe, especially with respect to in-
florescence architecture.
METHODS
Based on the family-level analysis of Bradford
and Barnes (2001
phyletic, and taxon sampling was designed to max-
) Cunonieae are clearly mono-
imize the geographic, phylogenetic, and morpholog-
ical diversity within this clade that has been
1998).
One distinctive species endemic to Sulawesi, Wein-
elucidated by previous studies (Bradford,
mannia descombesiana, is missing; otherwise, sam-
pling is broad, including exemplars from 45 species
of Cunonieae (Table 1). Codieae may be the most
closely related tribe to Cunonieae, but I was unable
to obtain internal transcribed spacer (ITS) nrDNA
sequences of Codieae to use as outgroups. Instead,
TS trees were rooted using two species from an-
other closely related tribe, Caldcluvieae (Bradford
& Barnes, 2001). The trnL-trnF cpDNA (Taberlet
et al., 1991) data set uses several outgroup taxa
from closely related tribes.
Collections were made from native populations
and botanical gardens between 1994 and 1998.
Fresh leaves were dried in silica gel for DNA pres-
ervation. All DNA samples are vouchered with her-
barium specimens and were deposited at MO and
in the country of origin. Table 1 lists source and
voucher information of each herbarium specimen
and GenBank accession numbers for all DNA se-
quences. Detailed information is available for Brad-
the TROPICOS
((http://www.mobot.org)).
ford collections on database
I sequenced both the nuclear ITS region (Bald-
win, 1992) and two adjacent chloroplast loci, the
tral. intron and the intergenic spacer between the
tral 1991). The
ITS region was sequenced first, which helped es-
3’ exon and trnF (Taberlet et al.,
tablish likely monophyletic groups. A smaller set
of trnL-trnF sequences was obtained from a sub-
sample of each distinct lineage that was discerned
from ITS data. Standard methods were used to ex-
tract, amplify, and sequence DNA loci, and these
are described in Bradford and Barnes (2001).
BLAST ((http://www.ncbi.nlm.nih.gov/BLAST/))
comparisons were done to confirm that sequences
were of angiosperm origin and not from possible
fungal or other contaminants, and indeed similar
nucleotide sequences were of appropriate loci and
within the eudicot clade.
Standard ITS primers (Baldwin, 1992) did not
strongly amplify or produce clear sequences of ITS-
2 for many species in Weinmannia sections Leios-
permum and Inspersae, or in Pancheria, Geissois
and Caldeluvia. Hypothesizing that high G-C con-
tent was interfering with PCR amplification, I de-
signed alternative primers with higher annealing
temperatures based on published sequences of 26S
rDNA (Kuzoff et al., 1998) and 5.8 rDNA sequenc-
es from my own work. The new primers did yield
clear sequences as hoped, but preliminary cladistic
analysis using ITS-2 data alone resolved clades
with a mixture of ingroups and outgroups, a result
incongruent with ITS-1, chloroplast, and morpho-
logical data (see Bradford, 2000, chapter 1, for a
figure of these results). This suggested that the ITS-
2 region amplified using the new primers was not
orthologous to ITS-1 sequences obtained using the
standard (e.g., ITS4) primers. To test this, a Parti-
tion Homogeneity Test (PAUP*4.0b6a; Swofford,
2001) was used to compare the ITS-1 and ITS-2
data sets, and they were found to be significantly
0.002). Because
incongruent (500 replicates, P =
Volume 89, Number 4
2002
Bradford 493
Molecular Phylogenetics in Cunonieae
ITS-1 and ITS-2 are adjacent loci they should have
the same evolutionary history; significant incongru-
ence between them strongly suggests that non-or-
thologous loci were amplified. The ITS region is
known to contain multiple copies of the ribosomal
genes, as well as pseudogenes, and different PCR
conditions can preferentially amplify different par-
alogous loci (Buckler et al., 1997). Until the issue
of paralogy can be resolved, the decision was made
to exclude the ITS-2 data from analyses of organ-
ismal phylogeny; these sequences, however, are
available in GenBank (phylogenetic data set with
the range AF521255-AF521298)
Sequences were aligned by eye in Se-Al (Ram-
baut, 1995) and exported in a NEXUS format. In-
sertions and deletions were scored as binary char-
acters (e.g.. present or absent). Any regions with
ambiguous sequence or uncertain alignment were
ignored during analysis. Parsimony cladistic anal-
yses were implemented in PAUP*4.0b6a (Swofford.
2001). For all parsimony analyses, the following op-
tions were used: characters unweighted and unor-
dered, searches heuristic, starting trees obtained
via random stepwise addition, TBR branch swap-
ping. COLLAPSE option on, STEEPEST DE-
SCENT option off, MULTREES on. Support for
clades was estimated with bootstrap values (using
the “Fast” bootstrap option with 10,000 replicates
in PAUP) and decay values (using Autodecay: Er-
iksson, 1999).
The nuclear and chloroplast data sets were com-
bined after checking for compatibility using the
Partition Homogeneity PAUP. T
could not reject the null hypothesis that the data
sets represent the same evolutionary history (500
= 0.06
Test in ns test
replicates,
RESULTS
ITs-] ANALYSIS
The final ITS-1 data set included 48 sequences
from 47 species (45 in Cunonieae) and a matrix of
260 characters. The number of equally parsimoni-
ous trees found during heuristic searches made it
impossible to complete branch swapping. Several
searches were done using random taxon addition,
and each analysis found the same large tree island
with 268 steps = 0.70, RI = 0.86). The strict
consensus of this tree island shown in Figure | is
based on over 30, ООО equally parsimonious trees.
The ITS-1 data strongly support the sister-group
relationship between Vesselowskya and the rest of
Cunonieae (Fig. 1). Pancheria is monophyletic and
weakly placed as the sister to clades of Weinmannia
and Cunonia. Five traditionally recognized groups
are resolved as clades, including Cunonia, and four
sections of Weinmannia: Weinmannia, In-
spersae, Spicatae, and Fasciculatae. Species from
Weinmannia sect. Leiospermum form part of a large
polytomy. The two sections from Madagascar, sec-
are placed as
sects.
tion /nspersae and section Spicatae,
sister taxa to each other. Not all of these clades
have high “Fast” bootstrap values, however, and the
data give no support for relationships between these
arger clades.
trnL-F ANALYSIS
The trnL-F data set included 996 characters
from sequences representing 38 species (29 in Cu-
nonieae). Ingroup sampling was less intensive than
with the ITS data set, but included sufficient sam-
pling from all major ITS clades. The parsimony
analysis found 420 trees of 244 steps (CI = 0.73,
RI = 0.82) (Fig. 2).
On the strict
monophyletic, and Vesselowskya is placed as the
sister group to other Cunonieae. In contrast to ITS
results. Weinmannia sect. Weinmannia is the sister
taxon to a large clade containing Cunonia, Panch-
eria, and all other species of Weinmannia. Cunonia
capensis, from South Africa, groups with Pancheria
while other Cunonia are monophyletic and sister to
the remaining Weinmannia. Malagasy Weinmannia
(sects. Inspersae and Spicatae) form a clade, as do
sections Fasciculatae and Leiospermum. The inter-
nal topology of the cladogram has low bootstrap val-
consensus tree, Cunonieae are
ues.
COMBINED ANALYSIS
For the combined analysis, 29 taxa, including 27
from the Cunonieae, were sequenced for both ITS
and trn L-F. The final data set included 1254 char-
acters. The parsimony analysis found a single most
parsimonious tree of 388 steps (СІ = 0.74, RI
0.81) (Fig. 3).
The base of the tree is structured similar to the
ITS trees: Vesselowskya is basal to all Cunoniae,
and Pancheria is a sister taxon to the clade con-
taining Weinmannia and Cunonia. Although there
is strong support for the position of Vesselowskya,
no other internal branches have high “Fast” boot-
strap or decay values. High “Fast” ES and
decay values do support most of the commonly rec-
including Pancheria and all five
Cunonia does
ognized taxa,
Weinmannia sections. By contrast,
not form a clade in the “Fast” bootstrap consensus
tree and has a decay value of one.
86SS8PAV Þ6L66ZAV LPI66ZAV spue[sp Á191908 (ауа 'OW) 226 paofpvag 9400] 7AV "f. SISUIIDIDA pruupumiog
LOSS8PAV CSA) puez мәм (OW) 9£6 paofpvag Y 7] »souia2na тииршиә
96SS8PAV spue[s] Á191908 (d¥d ‘OW) £16 piofpvag “18104 *£) Dsopfrasvd тииошиә
C6CC9t. AV C6166Z4AV CP 166c4V вшорәүе”) MIN (QON 'OW) £S0S surgdog SU) x "usuo1g DUOJOYIP piuupDumnog
(ZZ) voq *qq umuaedsoiz] иоцәәѕ
(IST) 7] отииошитәц
6£9S8PAV €1Z667AV 09166c.1V eensny (MSN ON) 628 paofpoag MOT[9d Y вәше ^M "Y “Splajezoy visnuaa DAYSMO]ASSAA
(c) dueg néysnojossa|
$ SE9SBTAV ешорәег) MIN (NON ON) 819 pzofpvag шшперте) ортол DILIYIUD]
© 1£9€9t AV PiuOpa[e7) MON (QON ON) 029 p10fpmıg шшперте) xo su) y "usuoag sapioopai]iud DILIYIUD
б 9€9CQt.4 V ZIZ66ZAV 6€ 166c.1V вшорә[ег) мәк (NON ON) 019 peofpoig ‘uedweg хә "әт отету D1421|2ung
© СЄОСӨРАҮУ LIZ66ZAV 8S L66ZAV vruopaſp MƏN (NON ON) 209 pu0/pv1g чә рирмәрЗиә DILIYIUD Y
2 PEOCRPAV bluopoa,ſp MIN (NON ON) 919 ptofpoig "uedureq mssayunsg DILIYIUD|
9 = (08) SUN x "usuoig nuauyoungq
5 el €€£9c9r. JV eruopo[e7) MƏN (QON ON) S£9 paofppag suc) Ņ "uxuoig pjjayoynd DIUOUN)
os ZEOCBHAV BIuOpayey) MON (QON ON) 609 puo/pv1g [YOG рирјиош DIUOUN)
g 2 [€9S8PAV OIZ66ZAV ZST66ZAV етиорәүегу MƏN (QON 70И) 209 p4ofpvig sut) y `иЗиолң рур\уЧолээш Dıuoun?)
£z 089684V 60Z66ZAV 9€ L66ZAV CSA) euyy quinos (OW) S£2 p4ofpvig "| szsuadp? pruoun-)
6Z9SBPAV 8076674 V CC [66c.1V eruopo[e?) MON (QON ОИ) 219 psofpoig SHS y “UBUOIG IDSUD]DG piuoum)
8c9€9t AV LOZ66ZAV PSL66ZAV вшорәүг) MIN (QON ОИ) £19 pl0fpvıg “AYO suaqnaoaqm piuoun:)
(QZ) "| Pruoun;)
(O12) "8 ^H AVAINON AD
[POS8PAV L1Z66ZAV POL66ZAV spue[s| uowojog (JISH ОИ) Org pzofpvag JUNY 22192122 sisdoavardg
SZZO6TAV ZLI66ZAV ity (VANS ON) cge psofpoig "pos 214078 “yo v9]MM4
ШЕ (VAS `ОИ) 610¢ surydoy a1dsa][1) pquadns sI0SS124)
81Z66ZAV COL66ZAV ецедѕпу (HA ON) 658 paofpvag эпи 4 Dupnunuaq S108812)
PZZO6TAV [L166ZAV eruopo[e7) MƏN (ПОМ ON) 009 psofposg urune[[1n<) 40]0951p mpo)
EZZ66ZAV 02 L66ZAV ei[Ensny (MSN ON) 258 paofpvag swoipuy prjofia4os DUI021]]D7)
Ot9cor. AV 9126624Ұ £9166c.1V (erpeagsn y ) op?) (16191€ OH) "ws saung uoq ‘q n;vjoiupd pianpop]p-)
C1Z66Z4V c9166c.1V рчрүвәў Moy (A ОМ) 606 p:ojpvag чипу) “y Dijofisos DUDIIY
ÞIZ66ZAV I9166c4V Pre. sn y (THA ON) Erg peofpoig MUA cp 7D) Y Avzneg (C[[onjy 74) оғорлотира DUDY Y
SAANOHILNO
ISLI As quu р—ә qua (рәї®лцүпә) uonoa[o^) (uOXP] ш saroads Jo 13quinu) snug‘)
UIZLIO JANEN sardadg
пәм ӘР exeyqns 5и pue әвәшоипт) и! satoeds jo 19qunu ay1 jo sS3Je wusa Jūuə lmno ]s9q əy} *ssoisoid ul HOM pue SUOISIAQI ]u32294 uo posed “DIUUDUU1A4 jo suor]oos
3 pue *p1ouos *aearuoun?) “sdno133no Aq poziuez10 oie sie[dusxy “USAS aie a9uanbas yore Jo Joquinu yuegua‘) әці pue sutouanbas VN( ш pasn sie[duioxa 10} SISYINOA °[ AQEL
495
Bradford
Volume 89, Number 4
2002
Molecular Phylogenetics in Cunonieae
[I9C914 V CS aU) (OW) 8&2 p0/pv1g “ARTY DULLIASOYIALL] рииюшиләд\
OL9OC8PAV erquiopor) (100 ON) 752 pio[pvag 7] 2s0ju2u10] DIUUDUUIIAf
609€8T.4 V 66 L66EAV 9V 166cAV Sn me] (06282 VIA) "ws “waaay q MIO оир
POOC8HAV р|әпләпәд (LMOd ON) S&P pyofpvag € лои ‘ds рииюэшизәд\
809<8 AV eIquio[o7) (ТОЭ (OW) Sz pio[ppag DNY DIJOf IÓ A DIUUDUUI9Mf
109€9t4V PIAI[OQI (ZSA ON) gee puofppvag spel, PYIUDJUD]]2 DIUUDUIUIAA
909€8T.4V 86166c4V CV LOOTA V ida hara! (ZSA OW) EZS pofpvag Aqsny пЎирф AAA и
соосу eog (781 ON) ZFS pao/poag '"uolorp] DID]NIUND юииэшипәу\
(11) "] эютииршщшәдү uonoos
CZOCEPAV JROSPSEPR IA (NVI OW) €€9 paofppag z ou “ds DIUUDULUIA
LEOS8TAV 90c66c.1V ES LOOGA V леоѕевереру (МҮЛ, ON) 0882 42q102]0 y quu nzioquoma pruupDunen
9C9S8TAV с0с66с4Ү сс166с4У лвәвр#ере|ү (NVL OW) 4Е©9 рго/рргя чә$ хә "D SISUILIDISDID DU AUDA
(6) p1ojJpeag 77) "f хә їрїриләң apsiadsuy U00YI3S
PZOSBPAV IROSRSEPR N (NVL OW) 059 p0/pv1g Jaye DAYIDISOUI]S ртииюшиләд\
EZOCOPAV POZ66ZAV LSL66ZAV JIPOSRSE PE N (NVI ON) SIZ pl0fpmıg ipreudag umapznsimmzups тииршиәд
CCOSBTAV COC66cAV 0€166c41V 1еоѕеӣвреуі (МУЛ, OW) 282 199 шоуу 1oxeg wopfimunu инип
12968 A V JPISPBRPR N (NVI ON) 269 P0/pD1g piojpeag ^S "f TNA `$ Y sisuatofoapui piuunuuiam
0c9c€9v.1 V JPISPBRPR IA (NVI ON) €69 рор ipreudagp DUDILIQUINY DIUUDUUIIA,
619€9T AV IRISPSEPP N (NVI OIN) 689 pio[ppag "ny, Punitelod DIUUDULUIDAf
8196844V IRISPBR PRIA (NVL OW) cro p0/pv1g piojpeag °) f (1ріешәя) руп DIUUDUUIIA
(9g) p1ojpeaq ^7» "f ‘Xe Iparuleg 2D7D21dg uonoas
LI9S8TAV ifr] (VAAS d ON) £co€ sudo Хеле) "y mon DIUUDUAHOR
9[9C€9T.4 V ваши р лере (4AN d OI) 62€ piofppag piojpeag `7 ‘f у surydoy 7 ^H 2pun]3004 DIUUDUUIAA,
CI9cort AV O2UIO£ (NVS ON) 100€ surgdog uo(] (| Xo “US (Z) nouixpaf nniununiog
€I9€8T4V c0c66c4V ША АЙ epsutuoq PISAP[P N (JAM d ON) 826 paofppag uoq (Y хә "ws (1) naumpaf DIU DULL Ta
TI9S8VAV L0c66c4V St 1000d V SHOES (dISH ОИ) r18 Pl0/p»ıg4 "us *) “Y pnzixo LUDA —
GLOS8SVAV 00c66c4V LV I66cAV e1sAeJe W (d ‘NIY ON) [TOS suzydoy SUNG PEU piuupunnam
(61) surydoy *) ‘н Y pue[goog хә rpieurog 2D7p]n212$D,[ uonoos
€£09€8T AV frg (VANS d ON) [ros surydoy MOOS ыы DIUUDUU124f
c09€8T.1V spue[s] Á191908 (ауа ON) [26 p2ofppag эңе 209590 piuununiog
166684V SUOUIO[OG («ІЯ ON) ZE8 paofpvag (29661 “sunydoy jo) [ слой ‘ds DIUUDWUIUIAA
009€9T.4 V PIUOP3]P) MIN (NON ON) 229 p0/pv1g SU x "идиота 0701295 DIUUDUUIIA,
LOOS8VAV 61664V PV LO6CAV рче[вэи MON (AMV COW) 216 P20/p01g ey Se pay Oe DIUUDUUTA A
66SE8VAV 9616671 V tr LOOZAV BOWES (OIN) 008 Plofpmıg ARI "V. SISUaOWIDS шири
ISLI Aa que pro ud (pajean¡no) UOT] JO") (uoxe] ur saroads Jo ләдшпи) snud‘)
UIBLIO JANBN saınad¢
‘ponunuoy “р әде,
496
Annals of the
Missouri Botanical Garden
ununodsolo7
p— Weinmannia raiatensis
Weinmannia sp.nov. 1
(—— — Weinmannia dichotoma
(0 —— ——— Weinmannia serrata
p— —— Weinmannia parviflora
lL Weinmannia samoensis
p— — Weinmannia racemosa
(000 0————— MWeinmannia silvicola
ا
Weinmannia
Weinmannia —
Weinmannia sp. nov. 3
Weinmannia auriculata
Weinmannia tinctoria
Un
N
S833
3333
AAAA
zc 2 3
8383383
gs БЕ
Sis
3SES
ЖЕЕ
за >
LEES
€IUUEUIUI9 AA
Weinmannia trichosperma
S 33 =
3 3 Е
& & &
з = =
32222
& RR y
SRA
$ & 2% F
= = %
333*32
& &
ASS
99]9[DOIOS? y
Weinmannia arguta
Weinmannia bojeriana
E
Weinmanni
Weinmannia stenostachya
Weinmannia sp. nov. 2
Weinmannia madagascariensis
Weinmannia rutenbergii
oes1odsu]
73 Pancheria brunhessii
[— —— Caldcluvia paniculata
| — L—— Spiraeopsis celebica
Figure 1.
Outgroups
Strict consensus of >30,000 most parsimonious trees of ITS-1 ше псев. “Fast” bootstrap values аге
above branches. Each section of Weinmannia is labeled to the right of the tree
DISCUSSION
In several respects, the phylogeny of the Cunon-
iae resulting from the combined analysis of ITS and
trnL-F (Fig. 3) is highly congruent with previously
published studies using morphological characters
(Bradford, 1998). It differs primarily by showing
that Weinmannia sect.
within section Fasciculatae. Most significantly, the
Weinmannia is not nested
combined data support the monophyly of Panch-
eria, Cunonia (albeit weakly), and all five sections
of Weinmannia. This analysis also suggests that
Malagasy taxa form a clade. Although the combined
‚ “Fast”
bootstrap and decay analyses do not give much
data set does produce a highly resolved tree
support for internal nodes, indicating that relation-
ships between major clades within Cunonieae are
still poorly understood.
In contrast to the results of Bradford (1998), the
molecular data sets do not uphold the monophyly
Volume 89, Number 4
2002
Bradford 497
Molecular Phylogenetics in Cunonieae
98
63
59
Weinmannia bojeriana
= Weinmannia sp. nov. 2
Weinmannia minutiflora
Weinmannia dichotoma
Weinmannia racemosa
Weinmannia silvicola
Weinmannia raiatensis
шпшщләд$отәт
Weinmannia samoensis
Weinmannia clemensiae
Weinmannia —
Weinmannia richii
oeje[norose,]
Weinmannia sanguisugarum
Weinmannia louveliana
Weinmannia madagascariensis
Weinmannia rutenbergii
PRES
[. Cunonia capensis
р, h Н J, Ў,
| о | »
oes1odsur W oejeoidg
Weinmannia marojejyensis
h.
Cunonia balansae
Cunonia macrophylla
Pancheria hirsuta
Pancheria phylliraeoides
Weinmannia bangii
Weinmannia trichosperma
Weinmannia tinctoria
Weinmannia tomentosa
ттийешшәү\
Vesselowskya venusta
Callicoma serratifolia
Codia discolor
Pullea cf. glabra
Figure 2. Strict consensus of trnl-trnF trees. “Fast”
mannia is labeled. Cunonieae taxa are shown in bold type
of Weinmannia. Weinmannia sect. Weinmannia has
a very long branch and is placed as a sister group
to Cunonia and other Weinmannia (Fig. 3).
long branch and the short internal nodes make it
This
possible that these results are not dependable. To
test whether this data set could statistically reject
the hypothesis of a monophyletic Weinmannia, |
built a constraint tree in MacClade (Maddison &
Maddison, 1992) making Weinmannia monophylet-
ic and enforced this topology in PAUP while re-
analyzing the combined data set. A single tree of
390 steps was found, only one step more than the
tree found in the unconstrained analysis. A
Ackama paniculosa
84 Ackama rosifolia
Caldcluvia paniculata
loh
ZU PF
100 Geissois benthamii
Geissois superba
Á— values are above branches. Each section of Wein-
COXON
the most parsimonious tree with the monophyletic
Weinmannia tree (Templeton, 1983; Mason-Gamer
& Kellogg. 1996), and the null hypothesis of a
monophyle tic tac could not be rejected N
‚Т = 2. P = 0.56). It would therefore be pre-
mature to — Bosse paraphyletic based
signed-rank test was then used to compare
on this data.
Morphologically, Weinmannia is recognized eas-
ily by the presence of hairs on the seeds, which are
lacking in other Cunonieae. Also,
and Vesselowskya have winged seeds, but
Pancheria, Cu-
попа,
i To account for
wings are lacking in Weinmannia.
498
Annals of the
Missouri Botanical Garden
Weinmannia bojeriana
d Woi ia marpiei
Weinmannia dichotoma New Caledonia
Weinmannia raiatensis Society Islands
Weinmannia samoensis Samoa
Weinmannia racemosa New Zealand
ogroed qnos
Weinmannia silvicola New Zealand
Weinmannia sanguisugarum
Weinmannia sp. nov. 2
IeoseaepejA
Weinmannia madagascariensis
Weinmannia rutenbergii
Weinmannia clemensiae
Malesia
Weinmannia fraxinea
Cunonia macrophylla
^
Melanesia
Weinmannia richii
Cunonia atrorubens
nnnin
New Caledonia
Cunonia capensis — South Africa
Ў W, nn han gii
e 60
g 1 Tropical America
NS 81 Weinmannia tomentosa
96 $ A
1 4100 Weinmannia tinctoria = Mascarenes
19
100. 79
12 <
100 |?
11
Caldeliuyi
Pancheria phylliraeoides
m Weinmannia trichosperma — Temperate
South America
Pancheria engleriana
New Caledonia
Pancheria hirsuta
Vesselowskya venusta — Eastern Australia
paniculata = South America
—— 5 changes
Figure 3.
given above branches, de
the geographic occurrenc of
monoecious, usually with асе flowers.
diu uM be En single most parsimonious
T
below. The
values
these character states with a paraphyletic Wein-
mannia, two additional morphological steps are re-
quired: either seed hairs were gained twice and
seed wings were lost twice (once on each of the two
Weinmannia lineages), or a reversal of both char-
acters occurred in Cunonia (Fig. 4). Other char-
acters supporting the monophyly of Weinmannia
== celebica = Malesia
spec le 9. Spec ies plac ‘ed in bold type
Outgroups
tree from the combined analysis. “Fast” bootstrap values are
> clades corresponding to each section of Weinmannia are labeled, as is
have a dioecious breeding system; others are
have been found in micromorphological studies by
R. Barnes (see Bradford & Barnes, 2001) in which
multicellular hair bases were found in all sections
of Weinmannia, but not in every species. These
kinds of hairs have never been found in other gen-
era of Cunoniaceae.
Although combined molecular data provide little
Volume 89, Number 4
2002
Bradford 499
Molecular Phylogenetics in Cunonieae
Hairs on seeds
Wings absent
Weinmannia
Cunonia
Wings absent
Hairs on seeds
Weinmannia
Pancheria
Vesselowskya
Weinmannia
Wings present
No hairs on seeds
Hairs on seeds Cunonia
Wings absent
Weinmannia
Pancheria
Vesselowskya
Hairs on seeds
Wings absent
Weinmannia
Cunonia
Pancheria
Vesselowskya
Figure 4. Alternative hypotheses for relationships
— Cunonieae. —a & b. Weinmannia is pue
with respect to Cunonia. This involves 38
o cs ular data set,
ones. —4 'inmannia is mot
390 — in * с ambined molecular data set, and 2 more
morphological one
support for the cladistic relationships among
Pancheria, Cunonia, and Weinmannia sections,
new insights into relationships within genera and
sections have emerged.
The South African species Cunonia capensis was
recognized by Bradford (1998) as being morpholog-
ically similar to two very distinctive New Caledon-
ian species, C. macrophylla and C. schinziana. All
three species have larger flowers and fruits than
other Cunonia, and similar inflorescence architec-
tures composed of axillary pairs of stout racemes at
the ends of stems. This inflorescence architecture
is more reduced than the compound racemes typi-
found in other
cally Cunonia (Hoogland et al.,
1997). Morphological cladistic analyses. showed
this group of species to be a highly derived clade
within Cunonia. In contrast, the combined molec-
ular data make Cunonia capensis basal within Cu-
nonia, and C. macrophylla basal within New Ca-
ledonian Cunonia (Fig. 3).
ost Cunoniaceae, including Cunonieae, are
pollinated by small flying insects, especially bees.
The distinctive features of Cunonia capensis and C.
macrophylla may be due to their unusual pollina-
tion biology. Observations show their visitors in-
clude nectar-feeding birds that perch at the base of
the raceme (Hopkins, pers. obs.; Coates Palgrave,
1983), which makes sense considering their rela-
tively large flowers and simple. rigid inflorescence
structure. The combined cladogram (Fig. 3) sug-
gests that similarities between Cunonia capensis
and Cunonia macrophylla may be plesiomorphic.
Given that bird pollination is only known in one
other Cunoniaceae genus (Geissois), it seems un-
likely that bird-pollination was plesiomorphic dur-
ing the origin of Cunonia, although it may be for
the extant clade. Alternatively, bird pollination may
be convergent in Cunonia capensis and C. macro-
phylla, but testing these hypotheses for ancient lin-
eages is nearly impossible.
Although molecular systematics only gives weak
support to the monophyly of Cunonia, the genus is
morphologically distinct. At least two characters
are shared only by species of Cunonia: fruits that
have a circumbasal-acropetal dehiscence, a char-
acter unique in the family, and floral disks that are
adnate to the base of the ovary, unlike any other
Cunonieae (Bradford, ds Bradford & Barnes,
2001; Rozefelds et al., 2001)
The most — clade in the analyses is
This is the
that of Weinmannia sect. Weinmannia.
argest section in the genus and is disjunct between
the Americas, where over 70 species occur, and the
Mascarene Islands, where 2 species are endemic.
The Mascarene species are distinguished by being
but otherwise are similar to American
species (Bradford, 1998). Although the topology is
not strongly supported, Mascarene species, repre-
dioecious.
sented here by Weinmannia tinctoria, are nested
between W. trichosperma, from temperate forests of
South America, and a clade of species from Neo-
tropical. montane forests.
Weinmannia trichosperma is apparently a rem-
nant of a more ancient, temperate lineage within
Weinmannia sect. Weinmannia—a lineage that
macrofossils show may have once occurred in Tas-
mania as well (Carpenter & Buchanan, 1993). That
a derived tropical clade of the section is disjunct
between the neotropics and the Mascarenes sug-
gests that interchange across the Atlantic and In-
500 Annals of the
Missouri Botanical Garden
Y €————— Racemes ——— al d
4 XÊ . Y -
Р *
М, m"
N N. A T Y y Y
e "мү NTP * К TT
$ N Y Б
= ЖФ. > y 7
NL a М ` als 4 y Li
ES "d t H A for " at
ar > Ж ALI [LPS 5
= LZ i - :
Inflorescence Modules (IM)
Figur Two common inflorescence architectures in. Weinmannia are shown; left, section Leiospermum; right,
section Tec latae. Racemes, internodes, and meristems are organized in particular patterns to form higher order
units called Inflorescence Modules (IMs), and IMs are organized in particular patterns along the main stem to comprise
the Total Inflorescence.
dian Oceans has occurred more recently than in-
terchange between temperate and tropical America.
However, few molecular characters
support this unexpected relationship, additional
data are required to test it.
In Bradford (1998), Weinmannia sect. Fascicu-
latae was paraphyletic with respect to a highly de-
since only a
rived Weinmannia sect. Weinmannia. This view is
overturned by the molecular evidence. Weinmannia
sect. Fasciculatae has high bootstrap and decay
values (Figs. 2—4). Missing from this study, how-
ever, is sequence data for Weinmannia descombes-
iana, an unusual species placed in section Fasci-
culatae, but with some traits suggestive of section
Leiospermum (Hopkins, 1998b).
Weinmannia sect. Leiospermum is widely distrib-
uted in the South Pacific, from the Bismarcks to
the Marquesas, and from Rapa to New Zealand
(Hopkins, 1998a, 1998c: Hopkins & Florence,
1998). Bradford (1998) recognized three species
groups within the section: New Caledonian species,
New Zealand species, and other South Pacific spe-
cies. The two species from New Zealand, Weinman-
nia racemosa and W. silvicola, each have distinctive
and unique inflorescence architectures. South Pa-
cific species outside of New Caledonia are dioe-
cious. As a whole, New Caledonian species have
no obviously derived features; however, Weinman-
nia dichotoma aborts its terminal meristem at every
node. Although each individual data set is ambig-
uous, the combined analysis strongly supports the
monophyly of section Leiospermum and places the
ew Caledonian clade sister to clades from the
South Pacific and New Zealand (Fig. 3).
A well-supported phylogeny presents an oppor-
tunity to reevaluate some of
characters 1 discussed in previous studies (Brad-
~
the morphological
ford, 1998). This earlier work emphasized inflores-
cence architecture and heterotopy, and some back-
ground on this is warranted here. The inflorescence
in Cunonieae comprises nested sets of structures.
The most familiar of these are the flower-bearing
ost genera, but ball-shaped
2001). These flower-
bearing axes, along with internodes and meristems,
axes, raceme-like in m
in Pancheria (Rozefelds et al..
are typically arranged in repeated units | term In-
florescence Modules (IMs) (Fig. 5). The structure of
IMs and their arrangement is highly variable among
Cunonieae genera and Weinmannia sections (Fig.
6). This observation led to a system of coding in-
florescence architecture based on principles of po-
q .
Molecular systematics has confirmed the general
sitional homology (see Bradford, 1998, for details
perception from comparative morphology that inflo-
rescence evolution represents heterotopy, and that
these characters can be effectively coded for mor-
phological cladistic analyses. For example, phylo-
genetic support for some clades in the morpholog-
ical cladistic analysis was based mainly on these
characters. The monophyly of Weinmannia sect.
Leiospermum was supported by two characters of
the inflorescence: having a sequential arrangement
of metamers within an IM, and having the largest
IMs in the terminal
5, 6, 7).
Furthermore, improved support for clades within
—
i.e., acrotonic) position (Figs.
section Leiospermum can be used to make specific
statements about the pattern of heterotopic chang-
es. From the generalized inflorescence form of sec-
tion Leiospermum (see top and bottom left diagrams
Volume 89, Number 4
2002
radford 501
Molecular Phylogenetics in Cunonieae
LA
Leiospermum Inspersae Spicatae Fasciculatae Weinmannia Cunonia
Figure 6.
phylogeny. The phylogeny is based upon Figure 3. but
though
ities to both sections
cl ide 9 havi
were collapsed. Cunonia and Pancheria — fences are t
C. macrophylla and C. capensis are similar to section Боан
Fasciculatae and Leiospermum. For detailed illustrations of inflorescence divers
Pancheria Vesselowskya
The general breadth of inflorescence diversity in Weinmannia is illustrated and linked to taxonomy and
ng no bootstrap suppor rt and a decay value of only
є Ч ) ч r
see REM (1998, 2001), Hoogland et al. (1997), the publication series of Hoskins (1998), and ren felds et al.
(2001).
of Fig. 7). it is clear that species from New Zealand
(Weinmannia racemosa and W. silvicola) have de-
rived heterotopic changes in their inflorescences
(see right diagrams of Fig. 7). Weinmannia race-
mosa has regained vegetative growth beyond the
inflorescence with a terminal vegetative bud, a re-
versal to the plesiomorphic state found in other sec-
tions of Weinmannia (Fig. 6). This species has also
lost the development of lateral inflorescence mod-
ules. The other species from New Zealand. Wein-
тапта silvicola, has developmental asymmetry: al-
though it does not always produce lateral IMs, when
present, they develop only from one axillary bud.
Furthermore, Weinmannia silvicola is the only spe-
cies to produce sequential metamers within an IM
and then abort the apical meristem. Other species,
such as Weinmannia dichotoma from New Caledon-
ia (see left diagram in Fig. 7), abort the apical mer-
istem at the first metamer within an IM. Although
abortion of the apical meristem occurs as part of
normal variation among IMs within many Weinman-
nia plants in the South Pacific, the fixation of this
trait is apparently derived within Weinmannia di-
chotoma. (See also Hopkins, 1998a, 1998c; Hop-
1998.)
Without a more resolved phylogeny it is difficult
kins & Florence,
to re-evaluate other characters discussed previously
Bradford, 1998), Although the
—
such as dioecy.
strong phylogenetic hypothesis for Weinmannia
sect. Leiospermum indicates that dioecy evolved
once within this clade (Fig. 3). it is difficult to dis-
cern the general pattern of breeding system evo-
lution within Cunonieae (see also Bradford, 1998:
2001).
Rozefelds et al.,
CONCLUSIONS
Molecular systematies has enabled us to confi-
dently delineate some major lineages within Cu-
nonieae and provided sufficient resolution in some
Un-
fortunately, little is known still about how Cunon-
clades to re-examine inflorescence evolution.
ieae genera and Weinmannia sections relate, except
Until
a better phylogenetic hypothesis is available, it is
for the basal position of Vesselowskya (Fig. 6).
м
best to retain the genera as currently circum-
seribed. despite questions about the monophyly of
Weinmannia. Remaining to be addressed is char-
acter evolution for many traits that vary among gen-
era of Cunonieae and sections of Weinmannia, al-
though evolution within some genera and sections
has been clarified.
This work is the first in-depth айели to under-
stand the phylogeny of Cunonieae. While taxon
sampling was broad, the number of sequenced loci
was likely insufficient to resolve all nodes. Ques-
502
Annals of the
Missouri Botanical Garden
Bismarcks, Solomons, Vanuatu, Fiji, Samoa,
Cooks, Societies, Rapa, Marquesas
New Caledonia
Weinmannia dichotoma
e.g. Weinmannia serrata
Figure 7.
The number of species and location of each clade is li
tions about character evolution may best be studied
by comparing variation within species, sections,
and genera rather than at the tribal level. For ex-
ample, Malagasy Weinmannia have the richest va-
riety of inflorescence architecture, sympatric Cu-
nonia species have a diversity of floral coloration
and scents, and in many Weinmannia species from
Malesia and the South Pacific is “leaky”
with morphologically male, female, and bisexual
dioecy
flowers found within a single population or plant.
Most importantly, studies on the spatial ecology of
species are almost totally lacking and would be
useful for effective conservation measures.
Literature Cited
Angiosperm Phylogeny Group. 1998. An ordinal classifi-
cation for the families of flowe "ring plants. Ann. Missouri
ot. Gard. 85: 531-553
,. 1992. Phylogenetic utility of the internal
transcribed spacers of ribosomal DNA in plants: An
e.g. Weinmannia samoensis
sp abeled. Some common forms of inflorescence archite
shown, and the most unusual species, Weinmannia silvicola, W. racemosa, and W. dichotoma, are highlighted.
New Zealand
Weinmannia racemosa
Relationships among clades within Weinmannia sect. Leiospermum based on the analyses presented here.
} |
cture are
— from the Compositae. Molec, Phylogenet. Evol.
1:
Bradford. j. C. 1998. A analysis of species-
groups in Weinmannia (Cunoniaceae) based on mor-
К апа inflorescence architecture. Апп. Missouri
Bot. Gard. 85: 565—593.
"2000 . Phylogenetic Systematics of Cunoniaceae
(Oxalidales). with an Emphasis on Spec
nflorescence E volution in
Genera. $a
cladistic
. Thesis, Washington nd St. Louis.
. 200 The applic ation of a cladis
the КҮ e ‘ation and identification of lada (C u-
noniaceae) in o ar and the Comoro Islands. Ad-
ansonia, sér, З, : 3: 237-246.
& R. В 2001. Phylogenetics and clas-
sification of Cunoniaceae (Oxalidales) using chloroplast
DNA sequences and morphology. Syst. Bot. 26: 354—
385.
&I . Berry. 1998. Cunoniaceae. Pp. 462—469
in P. E Be чту, В. К. Holst & К. Yatski evych (vol. edi-
tors), Pos of me Venezuelan Guayana, Vol. 4. Missouri
Botanical — n — St.
& г. 2001.
notes on he йога of the M
ouis.
New taxa and nomenclatural
arojejy massif, Madagascar—
Volume 89, Number 4
Bradford 503
Molecular Phylogenetics in Cunonieae
V. Cunoniaceae: Weinmannia. Adansonia, sér. 3, 23:
о
Buckler,
— banal DN
— netic implications. с netics 145: 812—832.
Carpenter, R. J. & A. M. Buchanan. 1993. Oligocene
leaves, p and flowers of the Cunoniaceae from Ceth-
ana, Tasmania. — Syst. Bot. 6: 91-109.
Coates Palgrave, Trees of Southern
Struik Poblishen. Ca ape m wn.
Eriksson, T. 1999. AutoDecay Ver. 4.0. Computer program
distributed by the author. Bergius Foundation, Royal
Swedish Ac — of r iences. Stockholm.
A. Ippolito yi T. P. Holtsford. 1997. The
: Divergent paralogues and
Africa.
Greuter, W., J. McNeill. Barrie, H. M. Burdet,
De — TS. iuc j^ T Nicolson, P. C. Si
J. E. ‚Р. Trehar urland & D. L. Hawk-
sent (editors) 2000. Inte d Code of e al
menclature (Saint Louis Code). Regnum Veg. 138.
Guillaumin, A. 1940. Matériaux pour la Flore de
— LVL Revision des Cunoniacées.
s . France 87: 242-256.
тена 3 1999. Flora of Ecuador; 78. Cunoniaceae. De-
partment of Systematic Botany—Institute of Biological
> iences, University of Aarhus, Nordlandsvej 68. dk-
3240 Risskov, Denmark.
Ha R. D., J. Jérémie & H. C. F. Hopkins. 1997.
enre Cunonia (Cunoniaceae) en Nouvelle-Calédon-
De SC de cinq espéces nouvelles. Adansonia.
la Nou-
Bull.
3,
Hopkins, x C F. 1 998 a.
noniaceae) in Malesia and the Pac
and an account of the species of Western Malesia. the
Adansonia, sér. 3, 20
A revision of d (Cu-
fic 1. Introduction
Lesser Sundas and the Moluccas.
541.
1998b. A revision of Weinmannia (Cunoniaceae)
in Maleaia and the Pacific. 2. Sulawesi and the Phil-
— sér. 3. 20: 43-66.
1908 revision of E einmannia (С unoniaceae)
in — and the Pacific. Guinea, Solon
Islands, Vanuatu and Fiji, ge notes on the species «
Samoa, Rarotonga bog Caledonia and New Zealand.
Adansonia, sér. б 11-106.
& J. p A revison of Weinmannia
(Cunoniaceae) in Malesia and the Pacific. 4. The So-
ciety, Marquesas and Austral Islands. Adansonia, sér.
3, 20: 107—130.
ippine S.
Kuzoff, I J . D. E. Soltis, P. S. Soltis & E.
A. ибо — lo potential ы entire
S rDNA sequences in plants. Molec. Biol. Evol. 15:
251-263.
Maddison, W. . R. Maddison. 1992. MacC *
Analysis of — and Character Evolution, V 3
Sine r, Sunderland, Massachusetts.
Mason-Gamer, R. J. & E. A. Kellogg. 1990. Testing for
р ylogenetic conflict among molecular data sets in the
ribe Triticeae ги пеае). Syst. Biol. 45: 524—545.
Mane Р. 1993. Our knowledge of the flora of New Cale-
donia: — and diversity in relation to vegetation
types and substrates. Biodiversity Lett. 1: 72-81.
Rambaut, A. 1995. Se-Al,
— program distribute by the Кетем Oxford.
Rozefelds, A. C., Н. W. nes & B. Pellow. 2001. A new
species "and shi cal morphology of ا lowskya
175-192.
equence Alignment Program.
(Cunoniaceae). Austral. Syst. Bot. 14:
Swofford, D. L. 2001. PAUP*: Phylogenetic — Us-
ing Painon, Ver. 4.0b6a. Sinauer, Sunderland, Mas-
Uni-
Gielly, G. Pautou & J. Bouvet. 1991.
versal | s for amplification of three non- — re-
gions of eorla DNA. PI. Molec. Biol. 1105-
1109.
1983. Phylogenetic inference from re-
Temple ton, A. R
eavage sile maps with partic-
striction endonuclease c
ular reference to the evolution of humans and the apes.
244.
Evolution 37: 221—
APPENDIX 1
The p are correct orthographies (following Greu-
ter et al., JO, ICBN Art. 21.2) and original publications
of all five pep see of Weinmannia. Types are also desig-
naled here for two sections.
. Nat.,
T
10, 2: 997, 1005, 1367.
< Weir nmannia pinnata
Сайан. 18:
Welomannis Le;
1759, nom. cons
Simplic TM Bernat
l., sine typo.
тла sect.
289. 1963, nom. D
Weinmannia sect. ospermum (D. Don) Engl., Nat.
flanzenfam, З (2а). 101. 1891. Leiospermum D.
Don, Edinburgh New Philos. м 9: E 1830. TYPE:
Weinmannia racemosa L A Suppl. 227. 1781. (lec-
totype, designated by H Uu 19982: 21)
Weinmannia sect. Racdmasae Berman, Bot. Jahrb. Syst.
83: 132, 185. 1964.
—
Weinmannia sect. Fasciculatae Be mardi ex —
& H. C. Hopkins, Adansonia, sér. 3: 998.
с Пав fraxinea (D. Don) a
Sides sect. Inspersae Bernardi ex J. C.
sect. nov. Weinmannia
Bot. Jahrb. Syst. 83: 132,
sine typo. TYPE: — d DC
This section was originally published without a ons
Bradford,
spe n 1€ es
Weinmannia - sect. Spi
sect. nov. Weinmannia sect.
Jahsh Syst. 83: 132. 1964, nom.
T : Weinmannia bojeriana Tul.
шш published without a type species
catae Bernardi ex J. C. Bradford,
nicatae Bernall, Bot.
invali d.. sine typo
This section was
—
CONTRIBUCION A LA Sandra S. Aliscioni?
FILOGENIA DEL GÉNERO
PASPALUM (POACEAE:
PANICOIDEAE: PANICEAE)!
RESUMEN
El género Paspalum L. presenta alrededor de 330 especies y una amplia distribución geográfica, — ipalmente en
las regiones tropicales y subtropicales de América. Debido a la marcada variabilidad mo rfológica entre sus especies,
— autores intentaron dividir al género en subgéneros, secciones о grupos informales de especies. En el presente
rabajo se realizó un análisis cladístico del género Paspalum utilizand s anatómicos foliares y exomorlológic os
con la — de poner a prue ‘ba su origen monofilético y кегин ‘er relaciones filogenéticas entre las especies mas
representativas. Asimismo se analizó el grado de homoplasia de los caracteres con e D to de estimar su valor
diagnóstico y calidad informativa en la caracterización de los distintos grupos. Para el anális
47 especies pertenecientes a 27 grupos establecidos según Chase con algunas modific 'aciones propuestas | por Ае lla
et al. y Morrone et al. Se identificaron 36 caracteres exomorfológicos y anatómicos foliares. La matriz de datos fue
analizada bajo el criterio de parsimonia con pesos iguales y con — implicito de caracteres utilizando los programas
Nona y Pee-Wee respectivamente. Sobre la base de los caracteres utilizados en este análisis no pudo establecerse un
origen común para el género confirmándose que el mismo representa una asociación parafilética de especies, estrec
mente relacionadas con Axonopus P. Beauv. y Thrasya Kunth, géneros también pertenecientes a la tribu Paniceae.
3
ч
ABSTRACT
Paspalum L. has — ‘ly 330 species and a large geographic : distribution, mainly in tropical and subtropical
regions of America. Due to the morphological variability among the species, different authors have proposed dividing
the genus into ш sections, or informal groups. A cladistic ТИР of Paspalum was conducted using anatomical
foliar and оре characters to test its monophyly and its phylogenetic relationships among representative
species. The degree of character homoplasy was have kinks in order to estimate the diagnostic value of anatomical
cha racia rs — groups. For the cladistic analys ed species of 27 groups established by Chase, with modification
ble by Cialdella et al. and Morrone et к nd 36 exomorphologi ‘al and. anatomical foliar characters were
considered. The Yo ata matrix was analyzed with Nona oa Pee-Wee, parsimony programs using equal weights and implied
weighting. The present phylogenetic analysis confirms that Paspalum is a paraphyletic group and shows close relation-
ships with Axonopus and Thrasya in Paniceae.
Key words: Axonopus, Paniceae, Panicoideae, Paspalum, Poaceae, Thrasya.
El género Paspalum L. es uno de los más im- La importancia económica del género radica en
portantes dentro de la tribu Paniceae debido al que muchas de sus especies son consideradas ex-
elevado número de especies que presenta y a su celentes forrajeras naturales, como por ejemplo Р.
amplia distribución geográfica, habitando princi- notatum Fliiggé, “pasto horqueta” y P. dilatatum
palmente regiones tropicales y subtropicales de Poir., “pasto miel."
América, con pocas especies en África y Asia. La El género incluye hierbas perennes, rara vez
falta de un estudio taxonómico global para este gé- anuales, cespitosas o rastreras, estoloníferas o ri-
nero hace difícil estimar en la actualidad el número zomatosas. Se distingue de otros miembros de la
total de especies. Según Chase (1929) el mismo tribu por presentar inflorescencias con racimos es-
incluye aproximadamente 400 taxones; Clayton y piciformes unilaterales; articulación de la espigui-
Renvoize (1986) consideran que comprende alre- Ila en la base junto al pedicelo; espiguillas abaxia-
dedor de 330 especies. les, solitarias o apareadas, dispuestas en 2 a 4
! El presente trabajo conatituye parte de la Tesis Doctoral prese ада en la Fac 'ultad Es Ciencias ач Че 1а
Universidad Nacional de La Plata. Expreso mi — "imiento a mis directores, el Dr. F. Zuloaga y la . M. Arriaga
por todo el apoyo — y a mis compañeros
y lectura crítica del manus
? Instituto de Botánica Darwinion. Labardén 200-C.C. 22-B 1642 HYD San lsidro, Buenos Aires, Argentina.
saliscioni@darwin.edu.ar.
ANN. Missouni Bor. Garb. 89: 504—523. 2002.
\agesen, L. Giussani y O. Morrone por las sugerencias aportadas
Volume 89, Number 4
2002
Aliscioni : 505
Filogenia de Paspalum
hileras, dorsiventralmente comprimidas, plano-con-
vexas o cóncavo-convexas; gluma inferior usual-
mente ausente y antecio superior endurecido obo-
voide a elipsoide (Chase, 1929; Burkart, 1969;
1990).
Debido al elevado nümero de especies y a la
Judziewicz,
marcada variabilidad morfológica entre las mismas,
diversos autores han intentado dividir a Paspalum
en subgéneros, secciones o grupos informales, de-
tectándose dificultades para dicha delimitación.
Estas categorías fueron establecidas sobre la base
de caracteres exomorfológicos, principalmente a ni-
vel de la espiguilla (pilosidad de la espiguilla, mor-
fología de la gluma superior, textura y coloración
del antecio, etc.) sin considerarse en la mayoría de
los casos los caracteres anatómicos foliares. Los es-
tudios taxonómicos llevados a cabo hasta el mo-
mento son fragmentarios: el género Paspalum ha
sido tratado en revisiones parciales y floras regio-
nales (Chase, 1927; Parodi, 1928, 1932, 1937:
Hitchcock, 1951; 1954, 1957, 1966, 1967:
Burkart, 1969; Rosengurtt et al., 1970; Quarín.
1975; Sendulsky & Burman, 1978, 1980a. b: da
Silva et al.. 1979: Pohl, 1980; Smith et al., 1982:
Renvoize, 1984, 1987a: Crins, 1991; Judziewicz.
1990; Filgueiras, 1993; Cialdella et al.. 1995: Mo-
1995, 1996, 2000; Rodriguez, 1998;
2000). Cabe destacar el valioso
Barreto,
rrone et al.,
Rodriguez et al.,
aporte realizado por Chase (1929) al estudiar las
especies de Paspalum que crecen en América del
Norte, estableciendo 27 grupos de especies con un
total de 140 taxones distribuidos en los mismos.
La anatomía foliar de Paspalum fue estudiada
por Tiirpe (1967), quien describió detalladamente
las características histofoliares de las especies ar-
gentinas sin aportar datos que permitan inferir re-
Arriaga (1998) lle-
varon a cabo un estudio anatómico comparado en
laciones entre ellas. Aliscioni y
los grupos Quadrifaria y Virgata del género Pas-
palum, destacando la semejanza anatómica entre
ambos. Posteriormente, Aliscioni (1999) realizó un
estudio histofoliar de las especies más representa-
tivas del género, observando que Paspalum es ana-
tómicamente heterogéneo presentando caracteres
diferenciales a nivel de especie o grupo de espe-
cies, algunos de los cuales mantienen cierta rela-
ción con los agrupamientos establecidos sobre la
base de caracteres exomorfológicos. Por otro lado
la autora muestra que los únicos caracteres anató-
micos comunes a todas las especies estudiadas son
aquellos asociados al tipo de vía fotosintética, los
cuales permiten corroborar que el género presenta
anatomía Kranz del subtipo MS, fisiológicamente C,
NADP-me.
Tiirpe (1967) define a Paspalum como un género
anatómicamente homogéneo y considera que los ca-
racteres diferenciales que se pueden observar en
las hojas, reflejan las condiciones ecológicas del
medio en el cual habitan las especies. Sin embargo,
Aliscioni (2000) estudió la anatomía ecológica de
algunas especies de Paspalum y describió diferen-
tes estrategias adaptativas presentes en el género
destacando que especies de hábitats muy diferentes
pueden manifestar características anatómicas si-
milares.
2
Existen estudios filogenéticos realizados en 1
tribu Paniceae los cuales incluyen un escaso nú-
mero de especies de Paspalum. Sobre la base de
datos morfológicos, Zuloaga et al. (2000) no logra-
ron resolver las relaciones dentro del género, mien-
(2001) y Giussani et al.
2001) realizaron análisis moleculares concluyendo
tras que Duvall et al.
—
en ambos casos que Paspalum sería parafilético.
En el presente trabajo se realizó un análisis fi-
logenético con el objeto de poner a prueba la mo-
nofilia de Paspalum. Se plantea una hipótesis sobre
a historia evolutiva del género estableciendo po-
sibles relaciones filogenéticas entre las especies
más representativas. Asimismo se analiza el valor
diagnóstico de los caracteres exomorfológicos y
anatómicos a la luz de las relaciones filogenéticas,
comparando el grado de homoplasia presente en
cada uno de ellos con el objeto de evaluar su im-
portancia en la caracterización de los distintos gru-
pos de especies.
MATERIALES Y MÉTODOS
SELECCIÓN DEL GRUPO INTERNO
El análisis cladístico se realizó sobre la base de
47 especies correspondientes a 27 grupos infor-
males de Paspalum establecidos según Chase
(1929, inéd.) con modificaciones propuestas por
Cialdella et al. (1995) para el grupo Bonplandiana
y por Morrone et al. (1995, 1996) en los grupos
Racemosa y Dissecta. Para seleccionar los taxones
terminales se consideraron representantes de la
mayoría de los grupos informales del género, tra-
tando de abarcar la máxima variabilidad morfoló-
gica como así también cubrir el área de distribu-
ción geográfica que presenta el género en América.
Se seleccionó una especie representativa de los
grupos Conjugata, Dissecta, Gardneriana, Inaequi-
valvia, Macrophylla, Orbiculata y Saccharoidea, y
dos especies para cada uno de los restantes grupos.
Cuando estuvieron disponibles, se seleccionaron
más de seis ejemplares por especie. El Apéndice
1 muestra la lista de ejemplares examinados.
506
Annals of the
Mou. Botanical Garden
SELECCIÓN DEL GRUPO EXTERNO
El grupo externo se seleccionó considerando ta-
xones que presentan un gran nümero de sinapo-
morfías potenciales compartidas con el grupo in-
terno (Nixon & Carpenter, 1993). Para ello se
seleccionaron representantes de géneros de la tribu
Paniceae los cuales presentan similitudes morfo-
lógicas y anatómicas con Paspalum (Burman, 1985;
Nicora & Rügolo de Agrasar, 1987; Watson & Dall-
witz, 1989; Crins, 1991; Morrone et al., 1993) y
que están relacionados filogenéticamente en análi-
sis realizados con datos moleculares (Gómez-Mar-
tínez & Culham, 2000; Duvall et al., 2001; Gius-
sani et al., 2001; Gómez-Martínez et al., inéd.).
Dichos géneros son Thrasya, Axonopus y Anthae-
nantiopsis Mez ex Pilger.
El género Thrasya fue relacionado por varios au-
tores con Paspalum, siendo ambos muy afines. Ni-
cora y Rügolo de Agrasar (1987) describen a Thra-
sya considerándola semejante al grupo Ceresia de
Paspalum por compartir la presencia de inflores-
cencias con racimos espiciformes solitarios y
membránaceo o folioso. Por otro lado, Thrasya se
diferencia de las especies de Ceresia por la apa-
y raquis
rente disposición uniseriada de las espiguillas y la
presencia de la lemma estéril hendida. A diferencia
de esto, Burman (1985) compara a Thrasya con los
grupos Dissecta y Decumbentes del género Pas-
palum refiriendose a ciertos caracteres de Thrasya,
como por ejemplo la lemma inferior hendida, un
'arácter que ha sido observado en especies del gru-
po Decumbentes, y la presencia de lemma inferior
delgada la cual acerca a Thrasya al grupo Dissecta.
Thrasya también comparte con Paspalum el mismo
tipo anatómico y fisiológico de vía fotosintética.
Giussani et al. (2001) y Duvall et al. (2001) reali-
zaron análisis cladísticos con datos moleculares,
mostrando la estrecha relación filogenética entre
ambos géneros. En el presente análisis se seleccio-
nó como — del género Thrasya a T. pas-
paloides Kunt
Watson y Dallwitz (1989) realizaron un análisis
fenético en el cual muestran a Axonopus y Paspa-
—
lum como géneros próximos. También Crins (1991
cita la relación fenética estrecha entre ambos.
Chase (1906) y Webster (1988) muestran que di-
chos géneros sólo se diferencian en la orientación
de la espiguilla en relación al raquis, siendo ad-
axial en Axonopus y abaxial en Paspalum. En re-
lación al tipo fotosintético, Axonopus es fisiológi-
camente С, NADP-me y
igual que Paspalum. Gómez-Martínez y Culham
(2000) realizaron estudios filogenéticos con datos
moleculares que confirman la cercana relación fi-
anatomicamente MS, al
logenética entre ambos géneros. En el presente
análisis se seleccionó como representante del gé-
nero Axonopus a А. compressus (Sw.) P. Beauv.
Morrone et al. (1993) realizaron la revisión del
género Anthaenantiopsis señalando que el mismo
presenta algunos caracteres morfológicos comunes
con Paspalum. Ambos géneros presentan especies
cespitosas, con inflorescencias contraídas a ligera-
mente abiertas, espiguillas pilosas solitarias o apa-
readas, antecio superior endurecido, piloso o gla-
bro, tipo fotosintético C,, subtipo anatómico MS y
ntimero cromosómico básico x = 10. Por otro lado,
Anthaenantiopsis se diferencia de Paspalum por
presentar inflorescencia espiciforme no unilateral;
espiguillas biconvexas, gluma, flor y pálea inferior
siempre presentes. Sobre la base de los caracteres
anatómicos foliares descriptos por Morrone et al.
(1993) se seleccionó A. rojasiana L. Parodi como
representante del género por presentar transcorte
foliar expandido al igual que la mayoría de las es-
pecies de Paspalum y haces vasculares con vaina
parenquimática externa remanente,
también se observó en P. inaequivalve (Aliscioni,
carácter que
El reconocimiento del tipo anatómico y fisioló-
gico asociado a la vía fotosintética constituye una
serie de caracteres valiosos para realizar clasifica-
ciones (Hattersley & Watson, 1992), especialmente
dentro de la tribu Paniceae la cual incluye géneros
con diferentes tipos fotosintéticos. Con la finalidad
de mejorar la resolución del grupo externo e im-
pedir incorrectas topologías del grupo interno (Nix-
on & Carpenter, 1993), se incluyeron como grupo
externo representantes de algunos de los tipos fo-
tosintéticos presentes en la tribu, diferentes al de
Paspalum. De esta manera se seleccionaron Erio-
chloa Kunth (tipo fotosintético C, РСК, subtipo
anatómico PS) y Panicum laxum Sw. (tipo fotosin-
tético C,).
A pesar de que Eriochloa comparte un menor
nümero de sinapomorfías con Paspalum, ambos gé-
neros presentan inflorescencias constituidas por ra-
cimos unilaterales, diferenciándose Eriochloa por
la presencia de espiguillas con un callo basal dis-
coide y número cromosómico básico x = 9 (Arriaga,
2000). Debido a que Eriochloa es un género mo-
nofilético y anatómicamente homogéneo (Arriaga &
Aliscioni, 2000) se seleccionó E. distachya Kunth
como representante del género.
Brown (1977) y Hattersley y Watson (1992) pro-
ponen distintos esquemas evolutivos para la tribu
Paniceae, pero en ambos casos sugieren a un gé-
nero anatómicamente non-Kranz como precursor de
la tribu entre los que incluyen representantes C,
de Panicum. Por tal motivo se asignó Panicun la-
Volume 89, Number 4
2002
Aliscioni
Filogenia de Paspalum
507
xum Sw. como raíz del árbol filogenético, ya que
dicha especie al igual que toda la sección Laxa del
subgénero Phanopyrum es fotosintétic — C
con anatomía non-Kranz (Zuloaga et al.,
3
CARACTERES
Se identificaron un total de 36 caracteres de los
cuales 14 corresponden a caracteres exomorfológi-
De los
16, 21, 23,
26,31 y 32) pennant a caracteres multiestado
cos y 22 a caracteres anatómicos foliares.
36 caracteres utilizados, 7 de ellos (3,
los cuales han sido tratados como no aditivos. La
lista de caracteres se incluye en el Apéndice 2 y
la matriz básica de datos en el Apéndice 3.
Las observaciones se realizaron sobre más de
300 ejemplares de herbario pertenecientes a las si-
guientes — iones: BA. — BAF, CAY, COL,
CPUN . FCQ, G, HUT, LPB, MEXU, MO,
NY, Р. P UNE SE, SL v USAM, USM, VEN,
segtin siglas tomadas de Holmgren et al. (1990).
Para la obtención de los caracteres anatómicos
se realizaron transcortes foliares y preparados epi-
dérmicos los cuales fueron observados mediante la
utilización de microscopia óptica. Se seleccionaron
caracteres histofoliares cualitativos que resultaron
ser consistentes para cada una de las especies es-
tudiadas, evitando incorporar aquellos que pudie-
sen modificarse segtin el grado de desarrollo de la
lámina. Los caracteres anatómicos fueron designa-
dos siguiendo la terminología propuesta por Ellis
(1976. 197‹
También se incluyeron todos los caracteres exo-
morfológicos diagnósticos de los grupos informales
de especies (Chase, 1929; Cialdella et al., 1995:
Morrone et al., 1995, 1996; Zuloaga & Morrone,
inéd.) correspondiendo principalmente a caracte-
rísticas de las inflorescencias y
J
las espiguillas.
ANÁLISIS FILOGENÉTICO
El análisis cladístico se realizó sobre la base del
1983).
Para ello se realizaron estrategias de búsqueda con
principio de máxima parsimonia (Farris,
pesos iguales y pesos implícitos de caracteres, uti-
1.8 (Goloboff.
8 (Goloboff, 1997b)
Los caracteres fueron pesados de
ands los programas ver.
y Pee-Wee ver res-
pectivamente.
acuerdo a su grado de homoplasia, asignando me-
nor valor a aquellos caracteres más homoplásicos
(Goloboff, 1993). La relación entre el peso asignado
a un carácter y su grado de homoplasia se expresa
mediante una función cóncava que incluye una
constante de concavidad (K) la cual puede variar
segtin la intensidad con la que el programa pena-
liza a los caracteres homoplásicos. Ramírez (1998)
sugiere que el valor óptimo de concavidad en un
análisis, sería matriz-dependiente y plantea que la
utilización de funciones moderadas de pesado (K
> 3) encuentra patrones jerárquicos más predicti-
vos. Goloboff (1995) considera que funciones de
pesado muy intensas (valores más bajos de K) no
deberían ser usadas, ya que en la práctica actúan
de manera similar a un análisis de clique. En el
presente análisis se discuten los resultados obte-
nidos con pesos iguales y con pesado implícito de
caracteres bajo el valor medio de constante de con-
cavidad (K — 3).
Debido al elevado nimero de taxones terminales
se optó por el método de búsqueda heurístico em-
pleando la opción mult* con 100 replicaciones la
cual genera árboles de Wagner con secuencias de
adición al azar en cada réplica, seguido de reaco-
modamientos del tipo TBR, reteniéndose 20 árboles
por cada réplica (h/20).
ron bajo la opción amb- la cual colapsa las ramas
Las búsquedas se realiza-
que están apoyadas solamente por optimizaciones
ambiguas. Las ramas de los árboles obtenidos fue-
ron posteriormente permutadas con el comando
max*. Con el objeto de sortear posibles islas de
árboles (Maddison, 1991) se utilizaron las opciones
Jump*1, jump*2 y jump*3 las cuales realizan rea-
comodamientos del tipo TBR en árboles con Та 3
pasos más largos respectivamente. Por último se
obtuvo el árbol de consenso estricto mediante la
opción nelsen (nel).
Como medida de apoyo para establecer el so-
porte de los diferentes clados se calculó el índice
de Bremer (BS) (Bremer, 1994) utilizando Nona ver.
1.8 (Goloboff, 1997a) para el análisis con pesos
iguales y Pee-Wee ver. 2.8 (Goloboff, 1997b) para
el análisis con pesos implícitos. Asimismo, se cal-
culó la frecuencia de grupo de Jackknife (Farris et
al., 1996) utilizando el archivo de instrucciones de
Nona/Pee-Wee jak.run, (con 1000 repeticiones, eli-
minando al azar el 36% de los caracteres de la
matriz y utilizando la opción de búsqueda mult*5)
y el programa fq.exe para calcular el árbol de con-
senso de mayoría (Goloboff, 1997a, b).
Para graficar los árboles y optimizar los carac-
teres se utilizó el programa Winclada Beta version
0.9.9 (Nixon, 1999). La optimización de caracteres
fue posteriormente verificada con el comando apo/
de los programas Nona (Goloboff, 1997a) y Pee-
Wee (Goloboff, 1997b).
RESULTADOS
El análisis cladístico realizado con pesos iguales
de caracteres (matriz analizada con Nona) encontró
2903 árboles, todos ellos igualmente parsimoniosos
508
Annals of the
Missouri Botanical Garden
= 0.29, Ri = 0.09),
Utilizando pesado implícito de caracteres, Pee- Wee
y de 143 pasos de longitud (Ci
encontró 28 árboles de 151 pasos de longitud (fit
1.2 y fit re-escalado = 43%). A continuación
se describen los resultados obtenidos del análisis
con Pee-Wee, mencionando las diferencias presen-
tes en el análisis de Nona.
Ambos análisis muestran que Paspalum se com-
porta como un grupo parafilético constituyendo un
clado de alto soporte (bs — 2) junto con Thrasya y
Axonopus (Fig. 1), sustentado por tres sinapomor-
fías: ausencia de gluma inferior (carácter que re-
vierte secundariamente en Thrasya); antecio coriá-
ceo (carácter que revierte secundariamente en P.
stellatum y P. polyphyllum del grupo Ceresia [ Pers. |
chb.; P. bertonii y P. lilloi del grupo Bertoniana
[sensu Chase, inéd.]; у P. saccharoides del grupo
Saccharoidea [sensu Chase, inéd.]); y ausencia de
vaina parenquimática rodeando los haces vascula-
res (Fig. 2A,
tituye el grupo hermano de dicho clado (Fi
B). Asimismo, Anthaenontiopsis cons-
En el análisis realizado con Nona, el clado con-
formado por Paspalum + Thrasya + Axonopus se
observa con un alto grado de colapsamiento; a di-
ferencia de ello el árbol de consenso obtenido a
partir del análisis realizado con Pee-Wee muestra
una topología más resuelta. Sin embargo ambos
análisis coinciden en que los tinicos grupos del gé-
nero Paspalum que constituyen clados monofiléti-
cos son Bertoniana (sensu Chase, inéd.) con bs —
1995) con bs
— 0.5, Anachyris Chase, con bs — 5 y Filiformia
1.3. Racemosa (sensu Morrone et al.,
—
(sensu Chase, 1929) con bs = 0.2; y que el grupo
1995) e
rafilético constituyendo un clado monofilético como
Bonplandiana (sensu Cialdella et al., s pa-
grupo hermano de Racemosa (bs = 0.8). Por otro
lado, los grupos Caespitosa (sensu Chase, 1929),
1929),
Eriantha (sensu Chase, inéd.), Fasciculata (sensu
Chase, 1929), Linearia (sensu Chase, 1929), Livida
(sensu Chase, 1929), Notata (sensu Chase, 1929),
Paniculata (sensu Chase, 1929), Parviflora (sensu
Chase, 1929), Plicatula (sensu Chase, 1929), Qua-
drifaria (sensu Barreto,
Ceresia, Corcovadensia (sensu Chase,
1966) y Virgata (sensu
Chase, 1929) son claramente polifiléticos o cons-
tituyen politomías no resueltas. El grupo Disticha
(sensu Chase, 1929) forma un clado monofilético
sólo en el análisis realizado con Nona, mientras que
el grupo Dilatata (sensu Chase, 1929) es monofi-
lético sólo en el análisis de Pee-Wee; aunque en
ambos casos sustentados por un bajo soporte.
A continuación se describen los grupos monofi-
léticos resultantes del análisis realizado con pesos
implícitos (Fig. 1) y las sinapomorfías que los sus-
tentan (Fig. 2A,
Paspalum lineare del grupo Linearia, junto con
las especies del grupo Filiformia se unen en un
2.5) por compartir la
presencia de porciones laterales del transcorte fo-
clado de alto soporte (bs =
liar inconspicuas o ausentes. Paspalum lineare es
el taxón hermano del grupo Filiformia, siendo este
último un grupo monofilético aunque de menor so-
porte (bs = 0.2), caracterizado por la presencia de
transcorte foliar de contorno semicircular (clado 1,
Fig. 1).
Paspalum proximun del grupo Linearia, es el
taxón hermano del grupo Bertoniana uniéndose en
un clado de bajo soporte (bs = 0.2) por compartir
la presencia de parénquima incoloro en las alas del
transcorte. Las especies del grupo Bertoniana for-
man un grupo monofilético de alto soporte (bs =
1.3) sustentado por la presencia de transcorte foliar
convoluto o subconvoluto (clado 2, Fig. 1).
Paspalum orbiculatum del grupo Orbiculata, se
comporta como taxón hermano del clado constituí-
do por las especies de los grupos Plicatula, Gard-
neriana y Virgata sustentado con alto soporte (bs =
1.3) por compartir la presencia de antecio obovoi-
de. Las especies de los grupos Plicatula, Gardne-
riana y Virgata se reúnen en un clado de alto so-
porte (bs = 1.8) por compartir la presencia de
antecio de coloración castaño (clado 3, Fig. 1).
Todas las especies de Paspalum consideradas de
los grupos Anachyris, Bonplandiana, Ceresia, Con-
jugata, Corcovadensia, Dissecta, Livida, Parviflora,
Racemosa y Saccharoidea, junto con una especie
de los grupos Eriantha (P. ammodes), Caespitosa (P.
y Quadrifaria (P.
coryphaeum), forman un clado en el cual se inclu-
—
indecorum), Notata (P. pumilum)
yen Thrasya paspaloides y Axonopus compressus (bs
— 0.2). Dentro de este gran clado puede destacarse
el nodo que retine a las especies de Racemosa y
Bonplandiana (bs — 0.8), sustentado por la presen-
cia de células distintivas Kranz, unido a P. repens
=
del grupo Dissecta como taxón hermano (bs =
(clado 4, Fig. 1). Las especies P. malacophyllum y
P. simplex del grupo Anachyris se reunen por l:
ausencia de la gluma superior en un clado de alto
—
М
soporte (bs = 5) (clado 5, Fig.
Los restantes nodos del árbol de consenso, a pe-
sar que algunos de ellos se muestran resueltos, no
poseen sinapomorfías que los sustenten como cla-
dos monofiléticos.
n relación al peso que recibieron los caracteres
1 función de la homoplasia, los caracteres exo-
Е #11 (nervios de la lemma superior),
13 (forma del antecio), 14 (coloración del antecio),
y anatómicos: #15 (porciones laterales del trans-
corte foliar), 24 (cloroplastos en la vaina mesto-
mática), 26 (vaina parenquimática), 27 (células dis-
Volume 89, Number 4 Alisc
2002
ioni 509
Filogenia de Paspalum
PANICUM LAXUM
RIOCHLOA
ANTHAENANTIOPSIS
Paspalum erianthum (Eriantha)
P. quadrifarium (Quadrifaria)
P. caespitosum (Caespitosa)
Р. inaequivalve (Inaequivalvia
—
P. urvillei (Dilatata)
P. fasciculatum (Fasciculata)
. notatum (Notata)
P. lineare (Linearia)
6 P. filiforme (Filiformia)
P. lindenianum (Filiformia
P. distichum (Disticha)
P. vaginatum (Disticha)
P. equitans (Fasciculata)
P. prox
imum (Linearia)
P. bertonii (Bertoniana)
P. lilloi (Bertoniana)
P. orbiculatum (Orbiculata)
P. modestum (Plicatula)
P. virgatum (Virgata)
P. plicatulum (Plicatula)
P. gardnerianum (Gardneriana)
P. regnellii (Virgata)
P sa haroides (Sa ha id
P. repens (Dissecta)
P. bonplandianum (Bonplandiana)
P. pallidum (Bonplandiana)
P. penicillatum (Racemosa)
P. pygmaeum (Racemosa)
P. stellatum (Ceresia)
THRASYA
P. lividum (Livida)
P. indecorum (Caespitosa)
. mandiocanum (Corcovadensia)
P. alcalinum (Livida
м
P. coryphaeum (Quadrifaria)
. malacophyllum (Anachyris)
P. simplex (Anachyris)
P. crispatum (Parviflora)
P. polyphyllum (Ceresia)
x
©
US
P. conjugatum (Conjugata)
P. parviflorum (Parviflora)
Figura 1. Arbol de consenso estricto obtenido a partir del análisis realizado con Pee-Wee (con pesado implícito de
caracteres). El número sobre las ramas indica el valor de frecuencia de grupo de jackknife y el número ubic * debajo
de las ramas indica el índice de Bremer. Los clados señalados (1-5) se discuten en el texto. Clado 1: P lineare
(Linearia) + grupo Filiformia. Clado 2: P. proximum (Linearia) + grupo Dorn Clado 3: P. orbiculatum — ulata)
+ grupo Plicatula + grupo Gardneriana + grupo Virgata. Clado 4: P. repens (Dissecta) + grupo Racemosa + grupo
Bonplandiana. Clado 5: grupo Anachyri
Missouri Botanical Garden
Annals of the
510
“SOx JuP[q sojn: ӘЛЛӘ UOD UE. р as soorspg[douiou S3139818 SO] А so1dou sojn: 919 uO2 UB: әри! 35 serpourodeurs se] enay el ua BIJSINW VS anb [9918 [9 ua
ap UILINVO sisajugaed 31]u2 SOpE.LI2. IUD $3J19]2EI8» A s9[oq.re so] sopo} Ud UILINVO stsojugied әлә SOPE.113: ӘПӘ ug]s2 OU anb SIIAJIPIBO :Ope]s9 [9 Р. әри! SEUIEI se] әр ofeqap оләшпи
[9 A ДӘ]ӘР.ГР [? B. ори sewel sel 91405 оләшпи IH "99M -994 uoo sopru3]qo sosoruouirsaed seul So[Oqie 8с SO] әр oun әр ose] oj В S319]2EIE2 so] әр uot. »eziuimd() “dv Fa engi J
V
V1VOMIA !әибәл y E 0 g - z ‘Bly ua епициод
VNVIY3NOYVO WnueveUpseb ¿E? ай ын
ЛЕ 2
VNVINOLY38
VINWANIT unulxoJd “y
v1vinolosv4 sueynbo y L 0 0 0
VLVLON unjejou A — (607 6
. c
VLVINOIOSW4 unje[noroseJ “y
ЧОДА
Boy
V1VIVIIO Jenn d
&jejejp ‘а
VTIAHdOWOVN unj|Kqdo428ul y
VLVWINIINVA 02
92 CL
VIA1VAIQO3VNI 8۸/eA!nbeeu! y
, ET |
VSOLIAS3VWI unsojidsoeo
- | 9c vC CC
VINWSINOWND unuejupenb y LE LE
VHLNVIY3 шпуиеиә uinjedseg be i
sisdonueuseyju
id гоо ооо
coRPOUT ZE Bl Л 129 t
шпхеј шпдшед
511
Aliscioni
Volume 89, Number 4
2002
Filogenia de Paspalum
'uor)enumnuo^) "c 210814
3 q
VeOTSIANWd штошмеа y "2 Un 0
v1vonrNoo unjebníuoo ‘а F О,
‘eine eu 2
sndouox s
vIS3339 winijAyddjod y 21 г
YYOTHIAYYd uinjedsuo y 0
cé
xejduns y 001) 0
SIHAHOVNV TOI O
uinjjkgdooejeui y m 8
VIYV3IYOVNO uneeud/oo y TS
L L?
VQIAI шпиеэје y 1 EF Z
WOON чедо nini d O ce 0 0
— gy
шпигәориеш y 0
VISN3QVAO2HOO [e
8SUOp0A02100 ч F Lt
eÁses ae ZG
= BE x gg Z
VATION N Peete
VIS3339 шп]}е/ә}$ `
S = 18/1978 “y AMARA!
uneewbÁd y |
VSOW3OVMH [3
uinjejioiued y v
uinpi|jed
VNVIONY IdNOS8 UN
wnuelpuejduog y
VLOASSIO suede! y jd LLL 0
VLVLON unjund y “(| i
VHLNVINS ѕәрошше y
VAGIONVHOOVS sepiojeuooes uinjedseg
V -z ‘B14 ap uoroenunuoo wv
512
Annals of the
Missouri Botanical Garden
tintivas Kranz), 30 (parénquima incoloro dispuesto
en las alas), 33 (macropelos adaxiales) y 34 (ma-
.
—
cropelos abaxiales) presentan peso = 10 (fit = 1
correspondiendo todos ellos a caracteres libres de
homoplasia. Por otro lado, los caracteres exomor-
fológicos: #1 (ciclo de la planta), 6 (espiguilla ab-
axial), 9 (gluma inferior), 10 (gluma superior), y
anatómicos: 420 (cavidades aeríferas) y P (c 'élulas
75). Por
último, el carácter anatómico: #18 wipes foliar
buliformes) presentan peso — 7.5 (fit —
abaxial con costillas y surcos diferenciados) pre-
6.0 (fit
tantes caracteres presentan peso = 5.0 (fit
= 0.6), mientras que los res-
= 0.5).
senta peso =
DISCUSIÓN Y CONCLUSIONES
Sobre la base de los caracteres utilizados en el
presente análisis, no pudo establecerse un origen
monofilético para el género Paspalum. Siguiendo el
método práctico propuesto por Farris (1974, 199]
para identificar clados polifiléticos o parafiléticos
sobre un árbol determinado, Paspalum se comporta
en este análisis como grupo parafilético.
Los resultados obtenidos concuerdan en parte
con otros estudios llevados a cabo por diversos au-
tores, realizados sobre la base de caracteres mor-
fológicos y moleculares, coincidiendo todos ellos en
interpretar a Paspalum como género no monofilé-
tico (Zuloaga et al., 2000; Duvall et al., 2001
2001; Gómez-Martínez et al., inéd.).
En un estudio amplio sobre la filogenia de la
; Giu-
ssani et al.,
tribu Paniceae realizado con datos morfológicos,
Zuloaga et al. (2000) incluyeron representantes de
diferentes grupos de Paspalum, los cuales no for-
maron un grupo monofilético por reunirse en una
politomía no resuelta junto con géneros afines. Por
Gómez-Martínez y Culham (2000) anali-
zaron filogenéticamente la tribu Paniceae sobre la
otro lado,
base de datos moleculares obtenidos a partir de la
secuenciación del gen de cloroplasto trnL-trnFk,
donde Paspalum es monofilético formando un clado
cuyo grupo hermano es Panicum obtusum; sin em-
bargo no se consideró a Thrasya en este estudio y
a relación con Axonopus no queda resuelta. Pos-
teriormente, Gómez-Martínez et al. (inéd.) sugieren
a condición parafilética de Paspalum por incluirse
en un clado sin resolución interna junto con An-
thaenantiopsis y Panicum obtusum.
Recientes estudios filogenéticos sobre la subfa-
milia Panicoideae, aportan mayor información so-
bre la historia evolutiva de Paspalum. Giussani et
al. (2001) realizaron un estudio sobre una matriz
molecular obtenida a partir de la secuenciación de
gen de cloroplasto ndhF. Estos autores incluyeron
varias especies de Paspalum concluyendo que el
mismo representa un grupo parafilético, con la in-
corporación dentro del mismo de Thrasya formando
entre ambos un clado monofilético de alto soporte.
Asimismo, Duvall et al. (2001) comprueban lo mis-
mo a partir de análisis realizados con el gen de
cloroplasto rpoC2.
A diferencia de los estudios mencionados ante-
riormente, Renvoize (1972) sugirió que Paspalum
sería polifilético por presentar caracteres exomor-
fológicos semejantes en especies de grupos que es-
tán sistemáticamente distanciados dentro del gé-
nero, aunque no probó su hipótesis realizando
estudios de filogenia.
Los resultados obtenidos en el presente estudio
muestran que solamente los grupos Anachyris, Ber-
toniana, Filiformia y Racemosa se resolvieron en
clados con relativo soporte indicando que los mis-
mos representarían asociaciones monofiléticas de
especies.
En relación a los datos utilizados, se detectaron
caracteres exomórfológicos y anatómicos libres de
homoplasia los cuales proveen evidencia para la
caracterización de los grupos. Dentro de los carac-
teres anatómicos, el desarrollo de las porciones la-
terales de la lámina, la presencia de células distin-
tivas Kranz, la presencia de parénquima incoloro
en las alas de la lámina y la presencia de macro-
pelos en ambas epidermis, podrían ser valiosos en
futuros estudios sistemáticos del género; algunos de
ellos junto con otros caracteres exomorfológicos,
definen grupos monofiléticos como el caso del clado
que reune a los grupos Bonplandiana y Racemosa,
en relación a las células distintivas Kranz. Dentro
de los caracteres exomorfológicos utilizados tradi-
cionalmente en la caracterización de los grupos de
especies, solamente la presencia de nervios cons-
picuos en la lemma superior, la forma y coloración
del antecio se manifestaron libres de homoplasia.
Contrariamente, otros caracteres exomorfológicos
como por ejemplo el hábito de la planta, el número
de racimos por inflorescencia, el contorno del ra-
quis de los racimos, la pilosidad de la espiguilla
entre otros, manifestaron un alto grado de homo-
plasia lo cual disminuye su valor informativo de-
biendo ser utilizados con cierta cautela en la de-
limitación taxonómica de los grupos.
Según Farris (1991) solamente los grupos mo-
nofiléticos tienen una existencia real por poseer
una historia evolutiva independiente; por tal motivo
los resultados obtenidos no sustentan la existencia
de Paspalum como grupo natural. Debido a la am-
plitud y a la diversidad observada dentro de este
género, el presente estudio muestra una aproxi-
mación hacia su historia filogenética y suma una
nueva evidencia a que el mismo debe ser reestruc-
Volume 89, Number 4
2002
Aliscioni
Filogenia de Paspalum
turado taxonómicamente, incluyendo a Thrasya y
Axonopus, aunque la inclusión de éste ültimo no es
consensuada por estudios de otra naturaleza.
Literatura Citada
Aliscioni, S. S. 1999, Estudio inr Comparado de
ro Paspalum L.
Panicoideae: Paniceae). Trabajo
ав en Fac. Сі. Nat. UNLP:
‚20
Especies Americanas del Gé (Poa-
ceae: de Tesis Doctoral
275.
. Anatomía ecológica
пуат D — ae:
13
8: 1874
&
a en especies del género
Panicoideae: Paniceae). Darviniii-
na rriaga. 1998. Estudio histofoliar compa-
rado de lok grupos Virgata y Quadrifaria del género Pas-
palum L. (Poaceae: Panicoideae: Paniceae). Candollea
93: 333-348.
Arriaga, M. 2000. Austral South American species of
Eriochloa. Pp. 141-148 en S. W. L. Jacobs & J. Everett
(editores). Gr : Systematics and Evolution. Vol. 2
CSIRO, Collingwood, Australia
& 5 Aliscioni. 2000. Estudio biogeográfico a
Eriochloa Kunth Aj ais eaea). Gayana Bot. 57 (supl.): :
Barreto, 1. L. 1954. Las especies afines a Paspalum vir-
gatum en s ‘eres ‘a del Sur. Revista Argent. Agron.
21: 125-14
. 1957, E s especies de Paspalum con dos racimos
HE wow en Rio Grande do Sul (Brasil). Revista Ar-
gent. 1. 24: 89-11
. Las especies afines a dide di quadrifa-
rium — ae) en la América del S
re y templado. Darwiniana 14: 13
——. 1907. As espécies de Paspalum (Gramineae)
— a Paspalum corcovadense do Rio G. Sul.
— — ultura do Rio Grande do Sul, Bol.
-13
— Kk. 1994. — h support and tree stability. Cla-
distics 10: 295—304.
Brow n, W. V. 1977. The — syndrome and its subtypes
ass systematics. Me ‚ Torrey Bot. Club 23: 1—97.
йш, 1969. P is 369—411 en A. Burkart
(editor), У llustrada En — (Argentina). Colecc
1. Inst. i^ э pee |
The genus Tray H. B. K. (Grami-
Миё, 4:
| Notes on genera wt Paniceae.
Biol. Soc. Wah 19: 183— —
1927
. Paspalum. En
E
der clima sub-
ээ.
cretaria
~
А P roc.
Hitchcock (editor),
The grasses of Ecuador, Ган. pee Bolivia. Contr. U.S.
Natl. He — 434—455
229. he — Meri 'an species of Paspalum.
Contr. U. S. Natl. Herb. 28: 1-310.
rédito. парат pl South America. Hitche
and C я Library, Bota
stitution, Washington,
Cialdella, A. ы О. Могод & F. О. Zuloaga.
visión de las especies del género Pa: ak
\ ly Department, анна E
1995. Re-
um (Poaceae:
Panic via Paniceae) Grupo Bonplandiana. Darwinia-
na 33: 5
. Vega. 1996. Estudios sobre la variación
амак d las espiguillas en т = la Tribu
Paniceae (Poaceae , и 34:
Clayton, W. D. € S. A. Renvoize. on
nium: — 's of ilie — Kew Bull.,
H.M.S.O., London
Crins, W. i р The genera of Paniceae (Gramineae:
enera Grami-
Mns Ser. XIII.
Pa-
nicoideae) i in n ои United States. J. Arnold
Supp pl. Ser. 1: 171-312
Davic >. Ў — Japan in the Poaceae. Pp.
143—155 en T. R. Soderstrom, К. EN Hilu, C. 5. Camp-
bell & worth (editores). Grass — matics
— Evolution, Smithsonian N Ны Press, Washing-
Dengler. N. G., R. E. Dengler & P. W. Hatte rsley. 1985.
Differing ontogenetic origins of PCR (“Kranz”) she —
in leaf blade >s of C, grasses (Poaceae). Amer. J. Bot.
204—302.
Duvall, M. R., J. D. Noll & A. H. Minn. 2001. Phyloge-
netics oí Paniceae (Poaceae). Amer. J. Bot. 88: 1988-
1992.
Ellis, R. 1976. A procedure for standardizi
leaf anatomy in the Poaceae, I. The leaf blade as viewed
in — section. Bothalia 12: бй 9.
— —— 1979. A procedure for stan
leaf antomy in the Poaceae, Il. he — rmis as seen
in surface Bothalia 12: 641-671.
Farris, J. S. 1974. dics он of paraphyly and
و — Zool. 23: 548—554.
. The — * of pityingenal tic analyses.
Pp. 2 en EN Platnick & V. Funk (editores), Advances
in C а 's, Vol. 2. Columbia Univ. Ne w York
. Hennig defined paraphyly. Cladistics 7:
1t1
у iew.
_
=)
Press,
—
——., V. A. Albert, M. Kallersjé, D. Lipscomb & A. (
Kluge. 1996. а. P SYM outperforms —
bor- joining, ^ ladistics 12: 99-124,
Filgueiras, 1993. аео and critical notes
um (Po ceae: Pa-
on some Brazilian species of Paspalu
niceae). Acta Amazon. 23: 147 E v
Giussani, L., J. Н. Cota-Sánchez, ). Zuloaga & E. A.
ellogg. 2001 A molecular — of the grass sub-
es Panicoideae cr aceae) E — origins of
C, photosynthesis. Amer. J. Bot
1993. Estimating — ler we ira dorini
Cladistics 9: 83—91.
)5. — and we дЕп: А reply to Turn-
er ба fani e. Cladistics 11: 91-104.
Та. Nona ver.
id New Yor
1997b. Pee-Wee ver. 2: 8. Computer program and
documentation, New York.
Gomez-Martinez, R. & А. Culham. 2000. Phylogeny of the
d — with boa on the tribe Pa-
ce from the trnL-F cpl
L. Jacol
E
Goloboff, P. A
tree searc Уе
1.8. — program and
uceae: E iden
136-1 39 . en S. W. obs & J. beet —
Gras Syste matics 2 Evolution, Vol. 2. CSIRO
Collingwood, с
Hattersley
“Matson, 1992. Diversification of pho-
1
te is p 38-116 en G. P. € артап (editor).
Grass Evolution and Domestication. Cambridge Univ.
Press, London.
Hite lico rock, A. S. 1951. Manual of the Grasses of the Unit-
ed States. Se er iss “ón. U.S.D.A. en Publ. pos
Туан n, P. K., N. rnett. 1990.
Holmgren & I
Index He — Part 1 I. The urbi x of the World.
Regnum Veg. 120.
Judziewiez, E. J. 1990. Flora of the Guianas. Series A
Phanerogams. Fasc. 8. Koeltz Scientific Books, Kónigs-
tein,
Kellogg. E. A. 2000. Molecular ~ Кз ergo ws evo-
— in the Andropogoneae. Pp W. L.
Jacobs & J. Everett (editors). Gras — 8 wd
Evolution, Vol. 2. CSIRO, Collingwood. Australia
514
Annals of the
Missouri Botanical Garden
Maddison, D. R. 1991.
multiple islands of most-parsimonious trees. Syst.
40: 315— 328.
е O., T. Filgueiras, Е. Zuloaga & J. Dubcovsky.
1993. Revision of Anthaenantiopsis (Poaceae: Panicoi-
deae: 53.
a eae). Syst. Bo
‚ F. Zuloaga & E. — 1995. Revisién del
REN Racemosa del género Paspalum (Poaceae: Pani-
coideae: Paniceae). Ann. Missouri Bot. Gard. 82: 82—
116
za & F. Zuloaga. 1996. Revisión de las
énero Paspalum 1. (Poaceae: Panicoi-
deae: — grupo Dissecta (s. str.). Candollea 51:
103-
espec ie 5 ^ 1
E De nham, S. Aliscioni & F. Zuloaga. 2000. Re-
даба de las especies de Paspalum (Paniceae), subgé-
nero ; — Candollea 55: 105-13. 5.
Nicora, E. & Z. Rúgolo de Agrasar. 1987. Los géneros de
ташы as de América Austral. Не misferio Sur, Buenos
Nixon. K. 1999. Winclada d ta) ver. 0.9.9. Published by
the author, ae a, New Yor
—— ё arpenter, 1993. On outgroups. Cladistics 9:
413—426.
Parodi, L.
Serie.
H. 1928. Gramíneas ся P n Argentina. П
di ig (Buenos Aires) 9
. 1932. Algunas gramíneas nuevas o críticas. Phy-
sis (Buenos Aires) 9: 12-24.
. 1937. Contribución al estudio de las Gramíneas
del género Paspalum de la Flora Uruguaya. Revista
us. La Plata, Secc. Bot. 1: 21 1-29 ).
Pohl, К. 1980. Gramineae. En W. Burger (editor), Flora
Costaricensis. Fieldiana, Bot. 4: 350—392.
Quarín, С. 1975. Notas E el género Paspalum (Gra-
mineae). eps 3: 10.
Ramírez, M . How жа should a weighting func-
tion be? Pp 2 en Hennig XVII. 17th Meeting of
the Willi Hennig Society. Sáo Paulo, Brasil.
T S. A. 1972. Studies in the Gramineae:
кут n 451-45 55
The — of Bahia.
ХХХ.
Royal Botanical
байам. p»
1987a. New grasses from Paraná, Brazil. Kew
Bull. a E -925.
37b. A survey of — As anatomy in grasses
XI. Pa niceae. Kew Bull. < 9-708.
ш z, Н. R. 1998. El — ro Ceresia (Pers.) Rei-
del género — L. (Gramineae) en Vene-
zue Kj Emstia 8: 7-5
, E. amac o de & L. Guevara. 20(
а sección Pani iflora Rodriguez del género Paspalum
a (G pe en Venezuela. Ernstia 10: —143.
— ngurtt, R. Arrillaga de Maffei & Ehe un re
de Artucio. pon Gramíneas Uruguayas. e rsidad
de la Re E a, Montevidec
Rua, G. H Weberling. 1995. Growth form and inflo-
rescence structure of Paspalum lL. (Poaceae, Paniceae):
Torres
=)
‘omparative morphological approach. Beitr. Biol.
Pune n 69: 363—431.
Sánchez, Arriaga. 1989. El síndrome Kranz en
Arundincllae (Panicoideae-Poaceae) de la flora Suda-
mericana, Parodiana 5: 2
Sendulsky, T. & A. G. Burman. 1978. Paspalum species
of the Serra do Cipó (1): A epa to the study of
the Brazilian Poaceae. Rev. Brasil. Bot. 1: 1-15.
. 19804. Paspalum species a the Serra do Cipó
61-277.
The discovery and importance ч
Zo
—
ID: A не to the — of the Brazilian Poa-
ceae. . Brasil. p 23-
l 980b. A new species r Paspalum (Gramineae)
from Brazil. А 32: 487-489.
Silva, T. S. da G. Burman & T. Sendulsky. 1979. Es-
pecies de Paspalum н da wy do Cardoso, Estado de
Sâo — he ie 11-2€
Smith, L. C. i ali & R. M. Klein. 1982.
— . 1, Paspalum—Zea, Pp. 910-1407 en
‚ Flora Ilustrada Catarinense. Itajaf,
Stebbins, G. L. 1982. Major trends of evolution in the
Poaceae and their possible significance. Pp. 3-36 en J.
R. Estes, R. J. Tyrl & J. N. Brunken (editores), Grasses
and Grasslands: a matics and Ecology. Univ.
Oklahoma Press, Norma
Türpe, A. M. 1967 [1966]. Histoasonomia o a especies
argentinas del км ro Paspalum. Lillo 72.
19€
Vegetti, A. C. . Análisis tipológico d la ilr 'scen-
cia en — (P oaceae). Kurtziana 19 —160.
Watson, L. & М. Dallwitz. 1982. — Grass Ge nera:
Anatomy, Miranda and Key Australian. Na-
tional Unive rsity, Research Sc hool of Biological Scienc-
anberra.
. 1989. Grass Genera of the World: Il-
vesc of € — ‘ters, Descriptions, Classification,
Interactive Identification, Information Retrieval. The
Australian National — rsity, Research School of Bi-
ological Sciences, Ca
Webster, R. 1988. ae “of He North American Paniceae
(Poaceae: Panicoideae). Sy: 6-609
Zuloaga, F. O., R. Ellis € revision
of Panicum subgenus Phanopyrum section Laxa (Poa-
ceae: Panicoideae: Paniceae). Ann. Missouri Bot. Gard.
79: 770-818.
O. Morrone & L. Giussani. 2000. A cladistic
analysis of the Paniceae: A p — th. Pp.
—135 en : W. L. Jacobs & J. arcc
Grasses! Systematics and Pur "Vol. 2 CS
Collingwood, peers
ÍNDICE DE NOMBRES CIENTÍFICOS
GRUPO INTERNO
—
Paspalum alcalinum Mez (LIVIDA)
Р. ammodes Trin. (ERIANTHA)
P. bertonii Hack. (BERTONIANA
P. bonplandianum Humb. & Bonpl. ex Flüggé (BON-
PLANDIANA)
P. caespitosum Flüggé (C o ү"
P. conjugatum Berg. (CONJU
P. corcovadense Raddi (CORC —
P. coryphaeum Trin, (QUADRIFARIA)
P. crispatum Hack. (PARVIFLORA)
P. dilatatum Poir. (DILATATA)
distichum L. (D om
P. equitans Mez (FASCICUI
P. fasciculatum Willd. ex Е (FASCICULATA)
P. filiforme Sw. (FILIFORMIA
Р. gardnerianum Nees (GARDNERIANA)
P inaequivalee Raddi (INAEQU IVALVIA)
j 17084)
P. lindenianum A. Rich. (FILIFORMIA)
: lineare Trin. (LINEARIA)
Volume 89, Number 4
2002
Aliscioni 515
Filogenia de Paspalum
P. lividum Trin. ex Schldl. (LIVIDA)
P. macrophyllum Kunth м
P. malacophyllum Trin. (ANA S
P. mandiocanum Trin. (CORC on ADENSIA)
P. modestum Mez (PLICATULA)
P. notatum Flüggé (NOTATA)
P. orbiculatum Poir. (ORBICULATA)
P. pallidum Kunth (BONPLANDIANA)
P. paniculatum L. (PANICULATA)
P. parviflorum Rhodé ex Ms — IFLORA)
F. penicillatum Hook. f. (R SA)
P. plicatulum Michx. (PL IC a М
F. polyphyllum Nees ex Trin. (CERESIA)
P. pygmaeum Hack. (R AC ЕМ
P. quadrifarium Lam. ZR Su ae
P. regnellii Mez (VIRGATA)
Р. repens Berg. (DISSECTA)
P. saccharoides Nees (SACC pia
P. simplex Morong ex Britton (ANACHY
P. stellatum Humb. & — ex Fliiggé * RESIA)
P. urvillei Steud. (DILATATA)
P. vaginatum Sw. (DIST Tera
P. virgatum L. (VIRGATA)
GRUPO EXTERNO
Anthaenantiopsis rojasiana L. Parodi
Axonopus compressus (Sw.) eauv.
Eriochloa distac E Kunth
Panicum laxum
Thrasya — Kunth
APENDICE 1. MATERIALES EXAMINADOS.
ESPECIES DEL GRUPO INTERNO
Paspalum subg. Anachyris Chase
۴ — ои Trin.
NTIN atamarca: Altivalle de las Gran-
р p rentz en (US). PARAGUAY. Central:
Asunción, Rojas 6732 (US).
2. Paspalum simplex. — ex Britton
PARAGUAY. Central: Asunción, Hartley & Rojas
54185 (US): Trinidad. Pate: & Rojas 7895 (US). a
10811 (US). Cordillera: Piribebuy, Rojas 13445 (US
Chaco: Bahía Negra, Rojas 13813 (US). Neembucú щй:
Estancia "Yacaré Pilar,” Rosengurtt B-5527 (US): Río
Pilcomayo, Morong 1583 (US).
Grupo Bertoniana (sensu Chase, inéd. )
1. Paspalum bertonii Hack.
(GENTINA. Misiones: Candelaria, en isla del
Río Paraná, — 988 (51): Iguazú, Salto Uruguay.
Hunziker 9941 (SI); —€— Porta 55 (SI); San Ig-
nacio, Cabrera et al. 28787 . Boelcke et al. 5393
(SI); Jardín América, Quarín D a BRASIL. Pa-
гапа: Guairá, Reitz & Klein 12139 (US
2. к lilloi Hac
ASIL, Prodi Salto Iguassá, Rambo 53619
Grupo Bonplandiana (sensu Cialdella et al., 1995)
l. Paspalum bonplandianum Humb. & Bonpl. ex Flüggé
ER
P |. Cajamarca: Cajamarca, Smith & Sánchez
Vega 7508 (MO, SI). Lambayeque: Ferreñafe, Sagás-
tegui et al. 12831 (HUT, SI). ECU = Carchi: Tul-
cán, Werff & Gudifio 11042 (MO, SI).
2. — к Kunth
CUAD mbabura: S hi, Peñafiel & Ta-
mayo 892 MEL QCNE. SI). I Cajamarca: Ca-
— а, Se p Cabanillas — & Guevara 643
(CPUN, SI); Cayayup, Ochoa 1580 сүл oe
Ferre 2s Llatas uiis 1920 (HUT, SI).
Grupo Caespitosa (sensu Chase, 1929)
1. Paspalum caespitosum Flüggé
BELICE. El Cayo: — Ridge, Lundell 6426
(SI, US). MEXICO. Las Ruinas de Coba, Tellez & Ca-
brera 2353a (MEXU, SI).
2. Paspalum — 'orum Mez
ARGENTINA. Corrientes: Santo Tomé, Krapovic-
kas et al. domi I). Misiones: Cainguás, Zuloaga &
Deginani 506 (SI); Candelaria, номда 's 1357 (SD): Santa
Ana. Montes 2334 (SI); Villa Venecia. Renvoize 3024
(М). PARAGUAY. Guairá: Iturbe. Montes 12739
(CTES).
—
'aspalum subg. Ceresia (Pers.) Rchb.
1. сү polyphyllum Nees ex Trin
OLIVIA. Santa Cruz: Sara, — 5385 (US).
a Mato Grosso: Tres Chase 10740
Gerais: Serra de Cipó, Archer et al. 4995
(US). Parana: Ponta Grossa, Swallen 8772 (US). Rio
anta Ca-
agoas,
Rojas 10209 (US).
Hassler 11561 (US).
^ Papam stellatum Humb. & Bonpl. ex Flügg
tGENTINA. Misiones: Posadas. — 594
— COLOMBIA. Magdalena: Santa Marta. Smith
(US). PARAGUAY. Amambay: Hassler 11058
E A Caaguazú: Estancia “Plate San José.” Rosen-
gurtt B 5883 (US). Central: Asunción, Rojas 11005
(US): Tarumandy, Schinini 6207 (US). Cordillera:
Cordillera de Altos, Fiebrig 664 (US).
=
Central: Ypacaraí. Lago Ypacaraí.
Grupo Conjugata (sensu Chase, 1929)
l. Paspalum conjugatum Berg.
RGENTINA. Jujuy: ong Grande, Taylor et al.
11377 (MO, SI). Misiones: Apóstoles, Cabrera &
Saénz 29087 (S1): ‘Candelaria Montes 2455 (SD; El Do-
ral. Belgrano, Hunziker et
6 (SD: — Mont 6950 (SI); Posadas, Bur-
(SD: 5 vier, Cabrera et al. 28659 (Sl).
BOLIVIA. Santa оа Sara, Steinbach 1910 (SI):
Warnes. Zabala 627 (SI)
Grupo Corcovadensia (sensu Chase, 1929)
1. Paspalum corcovadense Raddi
BRASIL. Rio de Janeiro: Leblón, ни 3377
(US). Rio Grande do Sul: Arth 1/02 (US). Santa
Catarina: ltajaí. Smith & Reitz 6076 (US). Sao Pau-
lo: Capital Chávaca dos Morrinhos, Pickel 5803 (US).
2. Paspalum mandiocanum Trin.
ASIL. Rio de Janeiro: Dist.
Froes 51-11331 (US). PARAGU AY. AI araná: Íra-
а, Montes 11029 (US). Central: A dde "n, Rojas
l. 3346 6 (US): Lago Ypacaraf, Hassler 12481 (US). Guai-
: Villarrica, Rosengurtt B-5929 (US). Itapúa: En-
— Black &
Annals of the
Missouri Botanical Garden
carnación, Rojas 11191 (US). Paraguarí: Caapucú,
Anderson 1151 (US); s.l., Rojas 1334611 (US
Grupo Dilatata (sensu Chase, 1929)
1. Paspalum уц Poir.
ARGENTINA. Buenos Aires: Bañados entre San
Fernando y Pac — Correa s.n (SD; La Plata, Cabrera
3420 (SD). San Luis: Pedernera, Rosa 48 (SI). Santa
Fe: — — 31 (SD. BRASIL. Rio Grande
s. Barreto 1601 (SI); São Gabriel, Barreto
(SI); Sao Simao, del Mato 12382 (SI). Santa Ca-
tarina: Agua Doce, Smith & Klein 13529 (SI). CUBA.
Pinar del Río: near Mariel, Ekman s.n. (US). URU-
GUAY. Montevideo: Atahualpa, Herter et al. 334 (SI).
2, de pres urvillei Steud.
BRASIL. Mato Grosso: vicinity of Dourados, Chase
10951 US. Paraná: Coloniba, Dombrowski 4391
^
a
—
(US). Santa Catarina: ltajaí, Klein 2749 (US). Sao
Paulo: Alto da Serra, Lemos 31504 (US). COLOMBIA.
ти Urrao, Pohl & Betancur 15504 (US). PAR-
GUAY. Guaira: Villarrica, Jorgensen 3535 (US), Jor-
gensen 3635 (US).
е Dissecta (sensu Morrone et al., 1996)
. Paspalum л ns Berg
GENTINA. Huson Aires: Delta del Paraná,
Burkart rhe (SD, Burkart 4343 — Chaco: Rio de
Oro, Holmbe rg s.n. (SI). Entre Ríos: La Paz, Pedersen
; Puerto La Paz, Duos et al. 26830 (SI).
Santa Fe: Garay, León & Fossati 922 (S1); San Jeró-
nimo, Morello 16873 (S1): Islas del Río Paraná. More llo
з (SI). PERU. Loreto: Maynas, Gentry
. VENEZL UE es A. Guárico: Camaguán. Gin el
s 4353 (SI, VEN).
33
Grupo Disticha (sensu Chase, 1929)
I. Paspalum distichum L.
BOLIVIA. Coe — Sl, tchcock
thichas, Hite ei 22882 (US
0, ADOR. vie hine hat
Н, | betw. Terrier Rouge and Fort Liberté,
Bartlett 17 183 (US). P iir AGUAY. Central: in regione
r 12452 (US)
lacus Ypacaraí, Hass
Paraguay, Pavetti & Rojas 9448 (US); in regione cursus
inferioris fluminis Pilcomayo, Mani 322 (US). PERÜ.
Piura: Hacienda Buenos Ап
PUE RTO RICO. — e: s.l..
2. ii ever vaginatum Sw.
BRA = Bz hia: —
i: Marapanim, Davidse et al. 17785 B (US)
Santa bs Smith
Anderson 668 (US).
Chase 6483 (US).
alenat-
Montabo, Hoff 5168 (C AY, Р, US).
ragı
bertad: Trujillo, Angulo T 0279 (US
Chorillos, Asplund 10906 (US).
Grupo Gardneriana (sensu Chase, 1929)
1 ш Mon rianum Nee
BRASIL. Goié ig ei Glaziou 22583 (
Glaziou z 586 (US). Maranhão:
Grajahú, Swallen 3652 (US). Piauí: s.l.,
22793
"Hite hcock 19908 (US). EC- -
Asplund 6571 (US).
Ji playas del Río
Belém & Peirheiro 2985
179
агі: Caacupú, Anderson 1125 (U S). Р E RÚ. La Li-
S). Lima:
Р); s.l.,
Barra do Corda to
Gardner 3507
(P). GUAYANA. Ebini Experimental Station, Harrison
1786 (US). COLOMBIA. Arauca: 10 km SW of Cravo
Norte, Blydnstein 1522 (US).
Grupo Eriantha (sensu Chase, inéd.)
1. Paspalum ammodes Trin.
BRASIL. Goiás: Serra do Caiapó, Irwin & Soder.
trom 7040 (SI). Paraná: Jaguarihyva, Dusén 16025
(G). São Paulo: San Jose dos Campos, Morello s.n.
(
2. Paspalum — Nees ex Tri
BRASIL. Goiás: s.l., Glaziou "22491 (С). Glaziou
22489 (G), Glaziou 22. 589 (G). Paraná: Turma 22, Du-
sén 15669 (G). Rio de Janeiro: Ы
(С). PARAGUAY. Amambay: in re gione — cur-
sus superioris fluminis а Hassler 11924 (6); Sierra
de Amambay, Lomadas Esperanza, — lores (G).
Grupo Fasciculata (sensu Chase, 1929)
l. Paspalum equitans Mez
. Corrientes: Santo Tomé, Quarín et
al. 2753 (CT ES, SI, US). PARAGUAY. Paraguarí:
Este de la Cordillera de Villa Rica-Marais, Balansa 87
Р).
2. Paspalum. jesse а Willd. ex Fliiggé
3RASIL. Pará: Boa Vista, Swallen 31651 (US). CO-
Pennell 3933 (NY, US);
S). Magdalena: Cos-
US). PARAGUAY.
LOMBIA. Bolívar: Magangue,
Island of Mompos, Curran 366 (U!
ta del Caribe, Dugand 5259 t OL,
Central: s.l., Morong 535 (U
Grupo Filiformia (sensu Chase, 1929)
1. Paspalum. — Sw.
CL r del Río: — 1. Ekman 946 (SI, US).
2. парат аа А. Кісі
BA. Pinar del Rio: Mariel. Ekman 947 (SI, US).
Grupo Inaequivalvia (sensu Chase, inéd.)
и — e Raddi
BOLIVIA. La Paz: Milluguaya, Caripata, Hitchcock
22672 (US). B RASIL . Mato Grosso do Sul: Corumbá
on Rio Paraguay. Chase 11133 (US). PARAGUAY.
Caacu s.l.. Archer et al. 4799 (US). Central: in
regione lacus Ypacaray, Hassler 12401 (US): s.l., Mo-
ron ). Cordillera Caacupe Ramírez 406
(US): in regione cursus inferioris — Pilcomayo.
Rojas 96 (US); s.l., Balansa 106 (
Grupo Linearia (sensu Chase, 1929)
1. Paspalum — Trin.
BRASIL. Ceara: s.l..
— — 22473 5 (G). Minas Gerais: s.l.,
M, Piauí: s.l.. Gardner 2975 (G, P). Sao
— 351 (P); s.l., Riedel 949 (G); s.l.,
Gardner 2979 (P). Goiás: Ca-
Wed-
2. Paspalum @ Mez
VASIL. Santa Catarina: — Luz, Smith &
Klein 13319 (P) Agua Doce, Smith & Klein 13653
(US); Fachinal dos Guedes, cami e Klein 14014 (US).
Paraná: dores Jonsson 307a (P). Sao Paulo:
e — 28535 (US); Estacás — bacs
Pickel 5520 (US); ж Campo Alegre, Black 51-
10994 M PARAGUAY. Guairá: Villa P. а, "Balan-
sa 69 (G).
Volume 89, Number 4
2002
Aliscioni 517
Filogenia de Paspalum
Grupo Livida (sensu Chase, 1929)
l. Paspalum alcalinum Mez
COLOMBIA. Cundinamar La Esperanza,
García Barriga 10009 (US). PARAG UAY. Alto Par-
aguay: Puerto Casado, Rojas aw (US); Cerro Galván,
Rojas ds (US). Chaco: . Casado, Rosengurtt B-
5663 (l s.l., oe В- — (US); Estancia Gus-
tafson, Sa 190 (US).
2. Paspalum lividum Trin, ex Schldl.
{GENTINA. Entre Ríos: Paraná, аа А 661 (Sl);
San José, Colón, Zucol 655 (SI). Santa ) de Julio,
Pensiero & Vegetti 2740 (SF); San Justo. poen 2736
(SF).
—
c
Grupo Macrophylla (sensu Chase, inéd.)
. Paspalum mac pne Kunth
COLOMBIA oquia: Fredania, Archer 505
(US); Coc arina, Lindi 14 (P). Cauca: La Manuelita,
Pittier 897 (US).
Grupo Notata (sensu Chase, 1929)
l. Paspalum notatum Fliiggé
)MBIA. Antioquia: Medellín, Archer 785
(US). Cauca: La Manuelita, Pittier 845 (US). Cundi-
namarca: Fusagasuga, Juzepczuk 5317 (US). El Valle:
Hacienda “El Trejo,” Hernando García 6412 (US).
— — iras, Fosbe rg 19240 (US); s.l., Dawe 854
(US) ‚емга: Jardín Botánico, Archer
1721 a ES Bernardino Rojas 1006 (US). Ñeem-
bucú: Estancia del Sr. Medina Pilar, Ramirez 131
(US). TT betw ios buy and Paraguarí, Jor-
gensen 3537 (US): pa 110 (US)
2, т сл 5
ASIL. Rio de Janeiro: s.l., Sellow s.n. (P); Co-
С hana leen 4343 (P); Jardín — Alta Boa
Cha: 3 (US). Santa Catarina: Blumenau,
bs 10. 366 em Florianopolis, Canasvieiras, Klein
al. 6478 (US).
ноя Orbiculata (sensu Chase, 1929)
. Paspalum orbiculatum Voir.
BOLIVIA. Pando: Madre de Dios, Nee 31353 (US).
COLOMBIA. Amazonas: Trapecio amazónico, Schul-
tes & Black 8508 (US); Boca Loreto—Yaco, Schultes &
Black 46-171 (US). Antioquia: alrededores del Rio
сс Gutiérrez & Barkley 17C178 (US). El
Valle: Río alema. Cuatrecasas 21118 (US). PAR. A-
GUAY. € al: Carapeg guá, Rojas 3350 (US); Arroyo
Tobaty, yee "4839 (US).
Grupo Paniculata (sensu Chase, 1929)
1. Paspalum juergensii Hack.
BOLIVIA. La Paz: Nor Yungas, Beck 9152 (US):
Sur PN Hitchcock 22627 (US). BRASIL.
Serra do Mar, Dusén 3624 (SI). Rio Grande do Sul:
I т 41248 (US): Inst. Agr. Sul-Pelotas- R.
- Sul. | € dp Botánico), da Costa Sacco 119
dei COLC utumayo: Paianza el Fabaño,
Gamta bise MS. ECUADOR. Tungurahua: Valley
of Pastaza River, Hitchcock 21842 (US). Hitchcock
21749 (US)
Parana:
7
2. Paspalum paniculatum
BRASIL. Bahia: as Pedane ‘Go, Chase 7868 (US).
COLOMBIA. Boya Valley of Rio Pomera, (US
2200414). Cauca: "n Tambo, Godfrey s.n. (US). Cun-
dinamarca: Estación Santana, Dugand & Jaramillo
3891 (US). El Valle: Cordillera Occidental, Cuatre-
casas 19685 (US). ECUADOR. Imbabura: Lita, Acos-
ta-Solts 12567 (Us 5). PARAG UAY. Amambay:
tiplanitie et declivibus “Sierra de Amambay,” Rojas
10134 (US), Rojas 10738 (US). Caazapá: Abai region
San Juan Nepomuceno, Rojas 5835 (US). TRINIDAD
TOBAGO. Port of Spain: s.l., Hitchcock 10036 (US).
in al-
Grupo Parviflora (sensu Chase, 1929)
l. Paspalum crispatum Hac Е
BRASIL. Mato Grosso: Santa Rita do Araguaya,
Chase 11844 (US), Chase 12051 (US). PARA( GUAY.
Amambay: in altiplanitie Sierra de Amambay, Hassler
11975 (US).
2. Paspalum parviflorum d ex Pligg
BRASIL. Amapá: Rio Pedreira, Pires & Cavalcante
52142 (US); Macapá, м ۳ & Fróes 51- 12300 (US);
betw. Rios Cuyubin and Flechal, Pires & —
5237 ) (US). Mato Grosso: entre Bards de
Pimienta Buena, Kuhlmann 1682 (US
zal. — 6489 b Pará: Missis Island. Goeldi
95 (Us ; Pombal, ná do Tapará, Black 52-15565
Geom 5116 (US). GUYANA FRANCE-
л Че тогу Leándre, Hoock 1136 (US);
— Le Prieur 5 (US).
LS
Grupo Plicatula (sensu Chase, 1929)
l. deco modestum Mez
¿NTINA. Corrientes Capital, — 608
men : SD: Itatí, Arrocera Rzepecki, Ahum
920 (CTES, SD; Solari, Curup Cuatiá, Vander shail
1393 (SI). Entre Rios: La Paz, Burkart & Bacigalupo
; Puerto Soto, Burkart & Bacigalupo 21093
(SI). PARAG UAY. Central: orillas de lago Ypacaraí,
Hicken 111 (SI); San Bernardino, Rojas 1111 (US).
2. Paspalum go wee Michx
BRASIL. Paraná: — Dusén 13491 (G). Rio
Grande do Sul: Bagé, Estancia Taruma, rem to 1600
(SI); Faz. Experimental de Criagao Ci Cruzes, Ba-
rreto 1377 (S1); Potrero № 24, Tora 1475 (SD; 12
km de Vacaría, — 1720 (CTES, SD; Passo Fundo,
Quarín 1718 (CTES, SI). Santa Catarina: Agua Doce,
Smith & Klein es (SI).
T
Grupo Quadrifaria (sensu Barreto, 1966)
l. Paspalum coryphaeum Trin.
BRASIL. Bahia: subida al pico das Almas, Zuloaga
et al. 4837 (SI); Br-242, Zuloaga et al. 4774 (SI). Mi-
nas Gerais: Morro da Gloria, Glaziou 16684 (P). Rio
de 00 iro: Nova Friburgo, "ue; 13328 (P); Pào de
sucar, Chase 8393 Ш (US).
2. — quadrifarium Lam.
RGENTINA. — entes: Curuzü- — d
ü, Ro-
1644 (SI). URUGUA tigas: Arroyo Gu
sengurtt 11328 dem Тош Río Yí, a 641
(US).
Grupo Racemosa (sensu Morrone et al., 1995)
l. Paspalum penicillatum Hook. f.
BOLIVIA. La Paz: Nor Yungas, Beck 9158 (SI,
A): 0.7 km W of Chuspipata, Solomon 13692
(MO, SI). PERU. La Libertad: Otuzco, Sagdstegui et
al. 11569 (HUT, SI). Lima: Chancay, Trovar 375 (SI,
518 Annals of the
Missouri Botanical Garden
USM); Lima, Lomas de Atocongo, Cerrate 832 (SI. loaga et al. 3319 (SI). Corrientes: 42 km E de Itu-
USM). zaingó, — et uL 2298 (SI).
2. Paspalum pygmaeum Hack.
ARGENTINA, Jujuy: Santa = Burkart & Tron-
coso 11670 (SD; s.l., Parodi 16452 (SI). Tucumán:
Taff, Burkart 5280 (SI). BO LIVIA. "bdo ni
— ollo, Beck et al. 18079 (LPB, SI). La Paz: In-
, Villavicencio 288 (LPB, SI); Murillo, 17 kn al
este de La Cumbre por el camino Unduavi, Solomon
18263 (LPB, MO, SI); Omasuyos, Tiquina, Beck 2923
(SI). PERU. Aneash: Bolognesi, joc 1375 (
USM). EUM Huancavelica, Patacancha, To-
var "du 55 (SI, USM). Junín: Huancayo, Tovar 2164
(US US
—
м
N
Grupo Saccharoidea (sensu Chase, inéd.)
l. — fe ee чаны Nees
3 a Paz: Quebrada de Zongo, к a
— Avila 1831 (S1). COLOMBIA. Meta: 10
de la carretera Villavicencio a Bogotá, Zuloaga 3955
(SI); Quindio, Monte negro, Zuloaga & Londoño 4186
(SD. PANAMÁ. € facer savannas near El Valle, Duke
& к. 6612 (51, US). Chiriqui: Fotuna Dam area,
Churchill & mete 6141 FONS SI). VENEZUE
— entre Piedra
=
>
Federal: « entre Caracas y | a — Burkart 16994 a a
(SI
sl).
Grupo Virgata (sensu Chase, 1929)
1. Paspalum — Mez
ARGENTIN . Misiones: Iguazú, Guaglianone et al.
995 (SI). BR ASIL. Minas Gerais: Viçosa, Chase 942
SI). Rio de Janeiro: Petrópolis, Glaziou 9049 (P).
PARAG UAY. Central: Asunción, Rojas 5800 (SI).
2. eco virgatum L.
GUYAN/ ree — Lamaha € —
Hitchcock 1655 52 — JAMA . Savana entre Ew
ton y Linstead, Hitchcock өз 1 TRINIDAD Y ro.
BAGO. Te — ‘Prune hfield, Eggers 5553 (Р).
XICO. Tabasco: Tenosique, Matuda 3519 MER
—
=
ESPECIES DEL GRUPO EXTERNO
Anthaenantiopsis rojasiana L. Pare
ARG ENTINA ;orrientes: Trans Tomé, Burkart
19628 (BAA, SD); Estancia Garruchos, Cabrera 11871
Misiones: Posadas, Parodi 4552 (BAA).
Axonopus ам (Sw.) P. Beauv.
ARGENTINA. Santa Fe: Arroyo Natulipiahué, Cas-
tellanos 19443 (BA). BOLIVIA. Santa Cruz: Sara,
Steinbach 1865 (BA), Steinbach 6 PARA-
;UAY. Alto Paraná: Puerto Bertoni, Bertoni 11313
(BA). URUGUAY. e Pau 11212 (BA).
Thrasya paspaloides K
BRASIL. Distrito Federal: Brasilia, Zuloaga s.n.
(SI).
Erioc bn distac hya Kunth
OL.
anta Cruz: Velasco, & Carrión
Rua
calcarea c ursus supe rioris fluminis Apa, Hasslerl 103.
(BAA, BAF, G)
Panicum — Sw.
ARGENTINA. Buenos Aires: Isla Martín García,
— 46626 (BAA). Chaco: Puerto Antequera, Zu-
APÉNDICE 2. LISTA DE CARACTERES.
CARACTERES
1. Ciclo de vida de la ‘sient (O) perenne; (1) anual.
La mayoría de las especies del género son perennes; a
s. Ejemplo de g son las espe-
cies del grupo Rac 'emosa D rone et al., 1995), como asf
también — especies de los grupos Parrilla y Pli-
catula (Chase, 1929). Según ; Rua y Weberling (1995) las
especies — ocupan habitats específicos como altas
montañas, márgenes de bosques montañosos o sabanas
con una marcada estación seca. Por otro lado, Morrone et
al. (1995) interpretan al cielo de vida anual como un ca-
rácter evolucionado dentro del género, basándose en los
criterios evolutivos propuestos por MM (1982), Wat-
son y Dallwitz (1982) y Davidse (1987).
2. Habito de la planta: (0) cespitosa; (1) no cespitosa.
Rua y Weberling (1995) desarrollaron distintos modelos
para Hus ribir las formas de crecimiento presentes en Pas-
palum; considerando que el hábito cespitoso es el más
miis en el género. Por otro lado, las
eden presentar tallos decumbentes, rastreros,
especies no ces-
pitosas pu
largamente rizomatosos o estoloníferos; pertenecen a gru-
pos restringidos dentro del género los cuales siete qe
bajo condiciones ambientales particulares.
3. Número — racimos en las — encias: (0) 1 racimo;
(1) 2 racimos; (2) más de 2 racimo
La inflorescencia de Paspalum ha sido descripta tra-
dic و 'nie como una panoja constituida por racimos
espiciformes insertos a lo largo de un eje. Vegetti (1987)
describe a la inflorescencia de Paspalum como de tipo
politélico debido a que los racimos o paro ‘ladios largos
te rminan en una espiguilla estéril o en una porción estéril
del raquis. Segtin Vegetti (1987) y Rua y Weberling (1995)
as difer ‘rencias estructurales de las inflorescencias dentro
in desarrollo variable de la zona
de Paspalum se deben a
ento de la zona internodal del eje principal
‹ Basado en estos conceptos, Vegetti
(1987) considera que las inflorescencias con 1 ó 2 racimos
serían más evolucionadas dentro del género
4. — del raquis de los racimos de las inflorescen-
s: (0) subtrigono; (1) expandido.
El raquis de los racimos prese nta un transcorte subtrí-
gone en la mayorfa de las
los racimos se presenta anchamente — con pro-
yecciones laterales hialinas o foliosas. Esto se observa en
as especies de los grupos Ceresia, o Racemosa y
Bonplandiana, como así también en algunas especies de
los grupos Lachnea, Gardneriana y Plicatula. Thrasya
bi т comparte con las « espec les anteriormente citadas
la presencia de inflorescencias con raquis expandido.
Espiguilla adaxial: (0) ausente; (1) presente.
аз espiguillas que se dispone n solitarias sobre el r
quis se definen con posición adaxial cuando el dorso de
Volume 89, Number 4
2002
Aliscioni 519
Filogenia de Paspalum
la gluma inferior se orienta contra el raquis del racimo.
ribu
la
Zuloaga el al. (2000) c onaiderati que la espiguilla adaxial
es homóloga a la espiguilla de posición terminal que co-
rresponde a una ramificación secundaria del raquis del
racimo
Espiguilla abaxial: (0) ausente; (1) presente.
Las espiguillas que se disponen solitarias sobre el ra-
quis se definen con posición abaxial cuando el dorso de
la gluma inferior se — opuesto al raquis del racimo.
Segün Zuloaga et al. en las especies de la tribu
Paniceae que presentan perde pareadas, la Ed
abaxial es homóloga a la espiguilla de posic ión axial,
cual se genera a partir de una ramificación terciaria de 1
raquis del racimo.
7. Número de hileras de espiguillas por racimo: (0) dos:
(1) cuatro
Los racimos presentan dos hileras de espiguillas cuan-
do las espiguillas son solitarias y cortamente pediceladas.
En este caso, es simple determinar la orientación de la
misma (Clayton & Renvoize, 1986), siendo adaxial en
Axonopus y abaxial en las especies de Paspalum con es-
piguillas solitarias. Sin embargo, dentro del género Pas-
palum algunas especies poseen espiguillas — ob-
servá nio cuatro hileras de espiguillas por racimo
nümero de hileras de espiguillas en cada racimo o para-
cladio largo es una consecuencia del grado de ramifica-
ción del eje de los paracladios cortos (Rua & Weberling,
1995). La presencia de espiguillas dispuestas en pares es
un carácter frecuente dentro de las Paniceae y se consi-
dera que es una condición primitiva siendo interpretada
como una sinapomorfía potencial de la tribu (Kellogg.
2000)
8. Pilosidad de la espiguilla: (0) glabra; (1) pilosa.
sencia de espiguillas provistas de pelos es pro-
Tiv pis una es — ion para la dispersion por ad-
hesión o a través del viento de la diáspora. Ejemplos de
esto son Paspalum conjugan el cual presenta espigui-
llas con pelos distrib
un delicado borde par altamente efectivo para la dis-
persion por adhesión: y P. saccharoides que crece en zonas
esenta espiguillas con Mos pelos se-
sc vales facili-
87). E
algo elevadas v pr
dons intercalados
tan la dipersion — el vie
grupo Ceresia se caracteriza por P presencia ү, es smod
llas — uamente pilosas, con pelos sedosos distribui-
samente sobre los márgenes; un patrón similar
=
©
—
&
=
=
к
palum, como Dilatata y Dissecta, la presencia de espigui-
llas pilosas es un carácter polimórfo dentro del grupo.
9. Gluma inferior: (0) ausente; (1) presente.
Uni de ae carac terística as cae a miel —
pigu —
bargo, muc * especies de Pu E n phos:
alguna espiguilla del racimo con ls gluma inferior par-
с ialmente desarrollada, en algunos ejempla-
s de P. distichum y р. аш (C hase, 1929).
omo ocur
10. Gluma superior: (0) ausente: (1) presente.
La ausencia de gluma superior es un carácter poco fre-
cuente dentro de Paspalum, siendo una característica dis
ntiva del subgénero Anachyris. A pesar de ello, —
especies de dicho subgénero poseen gluma jd con
diferentes grados de desarrollo (Morrone et al., ). Asi-
mismo, algunas especies pertenecientes a ш, grupos Ra-
cemosa, Gardneñana. Parviflora y Reimaria, también ca-
recen de gluma superior — & Morrone, inéd.).
Renvoize (1972) considera que 1 de la gluma
superior se pudo haber originado — ndientementé va-
rias veces dentro ^i género, posiblemente como una len-
dencia adaptativ
11. Nervios de la lemma superior: (0) conspicuos; (1) no
conspicuos.
1 presencia de lemma superior con nervios conspi-
cuamente marcados es característica. de
Anachyris; la misma presenta la cara abaxial sur
cinco nervios prominentes, sie к ésta una sinapomorfia
del subgénero (Morrone et al., 2000).
subgénero
12. Consistencia del antecio: (0) membrandceo; (1) cortá-
ceo.
Según Cialdella y Vega (1996) el ane cimiento y es-
pecialización del antecio superior en las Paniceae podría
=
—
=
ш.
V
Ф
—
—
c
22
ее
-l
—
led que much
lum, junte imente con otros géneros de la misma tribu pre-
sentan antecios de consistencia más dura debido a que
estarían adaptados a mecanismos de dispersión de la di-
ora a través de animales herbívoros. La dureza del an-
* io permitirſa que el mismo se mantenga viable atin des-
pués de atravesar el tracto digestivo del animal.
B
z Forma del antecio: (0) elipsoide; (1) obovoide.
. Coloración del antecio: (0) pajizo; (1) castaño.
CARACTERES ANATÓMICOS FOLIARES
. Porciones laterales del transcorte: (0) ausentes o in-
conspicuas; (1) desarrolladas.
La mayoría de las especies de Paspalum presentan ho-
jas planas o expandidas: sin embargo en alguno
F laterales de la lamina no desarrollan, originando ho-
as de aspecto filiforme las cuale
foliares reducidos a : zona de la costilla central de la
do:
lámina (Aliscioni, 200(
S Casos las
s presentan transcortes
—
T
16. Contorno del transcorte foliar: (0) expandido o en for-
nvoluto a
ma de “V” abiert JNA ct — ado; (2) co
subc onvoluto; (3) jo ; (4) semicircular (ver Alis-
cioni, 2000, fig. 3).
17. Superficie foliar — con costillas y surcos diferen-
ciados: (0) ausente: (1) presente.
18. Superfici ie н — con — y surcos diferen-
ciados: (0) ausente; (1) p
19. Costilla — (0) adas (1) —
20. Cavidades aeríferas: (0) ausentes: (1) p
Las cavidades aeríferas observadas en distintas be
cies de Paspalum se presentan en la costilla central de la
lámina; las mismas pueden presentar distinto grado de
desarrollo pero en todos los casos se originan a partir de
la lisis de células рае as i incoloras. La presencia
de dicho carácte s que habitan lugares
con exceso de — (Aliseioni, 2000, fig. 2).
er se aso
2). ri y alternancia de haces vasculares: (0)
un haz de 3° orden entre un haz de l° y uno de 2°
520
Annals of the
Missouri Botanical Garden
orden; (1) tres haces de 3° orden entre un haz de 1° y
uno de 2° orden; (2) sin distinción entre haces de 2° y
orden.
22. Pe osición de los haces vasculares: (0) todos los haces
dispuestos en el centro del transcorte: (1) haces de 1°
2° orden dispuestos en el centro del transc orte y haces
5 3° orden desplazados hacia la cara aba:
—
La posición de los haces vasculares en el transcorte y
su patrón de — ‘ión es un importante carácter diag-
nóstico (E llis, 1976); sin embargo en algunas especies la
proporción а eed de los diferentes haces varía desde la
jm media de la lámina hacia el margen. Con la finalidad
de « um observaciones fuesen comparables, los carac-
ie y 22 fueron — en la porcion media de cada
una de las semilámin:
23. Relación entre los grupos > células — y los
haces озен — 30 : (0) células buliformes
— soie buliformes sobre un
en; 2 ts buliformes sobre tres o más
haces de 3° orden
La ubicación de las células buliformes con respecto a
los haces vasculares es un carácter que se asoc la, en par-
te, al фа fotosintético de la especie. Según Renvoize
9679/9 1 los géneros de la tribu — fotosintéti-
camente С, М den subtipo MS, «
hana Thrasya у disi, tei las células bulifor-
mes se ubican siempre sobre haces vasculares menores
о Axonopus,
A diferencia de esto, las células buliformes se ubican en-
tre haces vasculares en Panicum laxum, especie fotosin-
téticamente C, y en Клос hloa, — que presenta tipo
anatómico- -fisiológic o C, PCK, subtipo PS.
24. sn mestomdtica: (0) sin — (1) con clo-
25. Vaina ООС (0) sin cloroplastos; (1) con
cloroplas
оз caracteres 24 y 25 hacen referencia a caracterís-
tic as anatómicas que permiten inferir la vía fotosintética
species. Para establec тег las —
sintéticos se siguió el criterio de Brow
971) y lar et al. (1985). Dichos autores inde
la vaina Kranz de las ыо. C, con subtipo anatómico
MS, a la vaina interna de las especies С, con subtipo
anatómico PS y a la vaina interna de las especies C, non-
Kranz. s son mestomáticas
por su origen ya que se ge ^ ran del mismo meristema que
A diferencia de ello, las «
En los tres casos estas vainas
origina al tejido vascular. spe-
cies C, anatómicamente PS, presentan la vaina Kranz en
»sición externa la cual se homologa a la vaina paren-
quimática de sea "e les Cy, pues en ambos casos estas
el meristema fundamental juntamente
con lis células инш as del mesofilo.
26. Vaina parenquimática: (0) ausente; (1) remanente; (2)
presente.
Morrone et al. (1993) ооа en el género Anthae-
nantiopsis la presencia de algunas células por fuera de la
vaina mestomática, de aspe * poen y con —
gruesas, con o sin cloroplastos especializados; dichas cé-
lulas son interpretadas por los autores como restos de una
vaina a a en proceso de reduc
Los caracteres 24, 25 y 26 son características anató-
nicas M MM a distintos procesos fotosintéticos y fue-
ron ооа con el objeto de resolver relaciones еп
los taxones exter
—
=
os, con lo cual se puede prevenir algu-
nas topologias erróneas del grupo interno causadas por
incorrectas resoluciones del grupo externo (Nixon & Car-
penter, 1993).
27. Células distintivas Kranz: (0) ausentes; (1) presentes.
Dichas células presentan carac terísticas ultraestructu-
por más de cuatro células clorenquimati as; las célula
distintivas se conectan a los haces vasculares vecinos por
— de venillas transversales (Sánchez & Arriaga,
1989). Segün Haneler y Watson (1992) podrían inter-
pretarse como reducciones extremas de haces vasculares
de menor orden. Su presencia dentro del género Paspalum
fue corroborada para los grupos Bonplandiana y Race-
mosa, A citadas por Cialdella et al. (1995) y
Morrone et al. (1995) respectivamente.
28. Células buliformes: (0) cubriendo casi toda la super-
ficie; (1) en grupos definidos.
=
Las células buliformes se presentan en la epidermis
adaxial y pueden formar grupos regulares bien definidos
con aspecto de abanico, dispuestos sobre uno a tres haces
terciarios. En otros casos ibuyen en casi toda la
excepto sobre los haces mayores, diferencián-
dose de las restantes células epidérmicas solamente por
ser de mayor tamaño
se distr
superficie,
29. Grupos de células buliformes asociadas a tejido paren-
quimático incoloro: (0) ausentes; (1) presentes.
30. Parénquima incoloro dispuesto en las alas: (0) ausen-
presente.
—
~
Este carácter hace — ‘ia н — чойо
asociado a
el parénquima incoloro asociado
as células les y la zona parenquimática de la
costilla central.
31. Pared tangencial externa de las células ы аў
adaxiales: (0) recta; (1) convexa; (2) papi
32. Pared Si d externa de las células ніта
abaxiales: (0) recta; (1) convexa; (2) papilos
Los caracteres 31 y 32 se refieren a las células epidér-
micas observadas en corte transversal. Las mismas difie-
ren en la forma que adquiere la pared tangencial externa,
pudiendo ser totalmente recta шан 0) o
queada o convexa dando a la «
(estado Se — célula papilosa (estado 2
mente cuando la pared tangencial externa se evagina no-
toriamente Bn una protuberancia digitiforme.
33. Macropelos adaxiales: (0) ausentes; (1) presentes.
34. Macropelos abaxiales: (0) ausentes; (1) presentes.
La presencia de macropelos es un carácter muy fre-
cuente en el género Paspalum; la mayoría de las especies
poseen macrope plos en una o ambas epidermis, pudiendo
ser largos y a
engrosadas, con la base bulbosa hundida en la ке ине
asociada en algunos casos a células epidérmicas sobreele-
vadas formando un cojín basal. Si bien los mismos pre-
sentan una marc A variación en cuanto a su longitud,
frecuencia, espesor de la pared y ángulo de emergencia,
sólo se nid ea aquellos aspectos más estables por
Volume 89, Number 4
2002
Aliscioni 521
Filogenia de Paspalum
lo cual tales características no fueron d incluidas en el aná-
lisis debido su variación intraespecífic
35. Asperezas adaxiales: (0) ausentes; (1) presentes.
Asperezas abaxiales: (0) ausentes; (1) presentes.
El término aspereza se refiere a células epidérmicas con
la pared tangenc ‘ial externa aguda; las mismas pueden di-
el tamaño re-
»yección exte
de la misma (Ellis, 1976). Dentro (A género Paspalum és se
observaron formas intermedias por al se reunió a am-
bos tipos dentro de un mismo carácter.
ferenciarse €
N
—
< %
ш Y
—
Annals of the
Missouri Botanical Garden
Apéndice 3. Matriz de datos. Referencias: | | indica carácter polimorfo; — indica carácter no comparable
1 2 3 4 5 6 1 8 9 10 11 12 13 14 15 16 17 18
Panicum laxum 0 1] 2 0 l 1 1 [01] l | 1 0 0 0 1 0 l 0
Eriochloa 0 0 0 0 1 0 0 1 l l 1 0 0 0 1 0 О 0
Anthaenantiopsis 0 0 2 0 l l l l l | 1 0 0 0 1 0 1 0
Axonopus 0 1] 2 0 l 0 0 0 0 l 1 1 0 0 l 1 0 0
Thrasya 0 0 0 l l l l l l l 1 — 0 0 1.0 0 0
P. malacophyllum 0 0 2 | l l l 0 0 0 0 1 0 0 | 0 0 0
2 simplex 0 0 2 l l 1 1 0 0 0 0 1 0 0 1 0 0 0
P. bertonii 0 0 1 l 0 | 0 l 0 | 1 0 0 0 1 2 l 0
P. lilloi 0 0 l 0 0 | 0 1 0 1 | 0 0 0 | 2 1 0
P. bonplandianum 0 | 2 | 0 | 0 0 0 l l 1 0 0 1 0 0 0
P. pallidum 0 1! 2 І 0 | 0 0 0 | 1 |] 0 0 1 0 0 0
P. caespitosum 0 0 2 0 l | 1 0 0 1 1 1 0 0 1 0 0 0
P. indecorum 0 0 2 0 l | |] 0 0 | 1 1 0 0 1 0 1 0
P. polyphyllum 0 0 [2] 0 1 1 1 1 0 1 1 о 0 ото 0 0
P. stellatum 0 0 [12] 1 0 | 0 l 0 | 1 0 0 0 1 0 1 0
р. conjugatum 0 | l 0 0 | 0 l 0 | 1 1| 0 0 1 0 0 0
P. corcovadense 0 0 2 0 l l l 0 0 1 1 1 0 0 1 0 0 0
P. mandiocanum 0 0 2 0 1 1 |] 0 0 1 1 1 0 0 1 0 0 0
P. dilatatum 0 0 2 0 1 | |] | 0 1 1 1 0 0 1 0 0 0
? urvillei 0 0 2 0 1 l l l 0 1 1 1 0 0 1 0 0 0
P. repens 0 | 2 l 0 1 0 | 0 | 1 1 0 0 1 0 1 1
P. distichum 0 1] 1 0 0 1 0 [OI] [Ol] 1 1 1 0 0 1 0 l l
P. vaginatum 0 1] 1 0 0 1 0 01] 0 111 0 0 1 0 l 0
P. gardnerianum 0 0 2 0 l | 1 [Ol] 0 0 1 1 | | 1 0 0 0
Р. ammodes 0 0 2 0 0 1 0 1 0 ] ] 1 0 0 |] 1 0 0
P. erianthum 0 0 2 0 l | | l 0 | 1 1 0 0 1 0 1 0
P. equitans 0 0 2 0 0 | 0 0 0 1 1 1 0 0 1.0 1 0
P. fasciculatum 0 | 2 0 0 | 0 0 [OI] 1 1 1 0 0 1 0 1 0
P. filiforme 0 0 0 0 0 1 0 0 0 ] 1 1 0 0 0 4 — —
P. lindnerianum 0 0 [l] 0 0 1 0 0 0 | 1 |! 0 0 0 4 س —
P. inaequivalve 0 1] 2 0 1 1 1 0 0 1 1 1 0 0 1 0 0 0
P. lineare 0 0 [12] 0 0 1 0 [01] 0 1 1 1 0 0 0 3 — —
P. proximum 0 0 [12] 0 0 1 0 [01] 0 1 1 1 0 0 1] l 0 0
P. alcalinum 0 0 2 0 1 1 1 [01] 0 1 1 |! 0 0 1 0 0 0
P. lividum 0 1 2 І 1 1 1 [01] 0 |o l 1| 0 0 1 0 0 0
Р. macrophyllum 0 0 2 0 l | ] 0 0 | 1 |] 0 0 | 0 0 0
P. notatum 0 | | 0 0 | 0 0 0 1 1 1 0 0 1 0 1 0
P. pumilum 0 0 l 0 0 1 0 0 [01] 1 l l 0 0 1 0 0 0
P. orbiculatum 0 | | 0 0 | 0 0 0 1 1 1 1 0 1 0 0 0
P. juergensii 0 0 2 0 l 1 1 [01] 0 | 1 1 0 0 1 0 0 0
P. paniculatum 0 0 2 0 1 l | [01] 0 l l l 0 0 1 0 0 0
P. crispatum 0 0 [12] 0 1 1 1 0 0 l l 1 0 0 1 0 0 0
P. parviflorum г O [I2] 0 0 | 0 0 0 l l 1 0 0 1 0 0 0
Р. modestum 0 | 2 0 0 1 0 [Ol] 0 l 1 |] 1 1 1 0 0 0
P. plicatulum 0 0 2 0 l 1 1 [Ol] 0 | l 1 l | 1 0 0 0
P. coryphaeum 0 |] 2 0 l l 1 [Ol] 0 | |] 1 [Ol] O 1 0 0 1
P. quadrifarium 0 0 2 0 l | 1 [0l] 0 1 1 1 [ol] O 1 0 l [01]
P. penicillatum l l 2 l 0 | 0 0 0 1 | l 0 0 1 0 0 0
P. pygmaeum 1| 0 2 1 0 1] 0 0 0 ! ! ! 0 0 ! о O 0
P. saccharoides 0 0 2 0 0 | 0 1 0 l 1 0 0 0 1 0 | 0
Р. regnellii 0 0 2 0 l | 1 [0l] 0 1 1 I] 1 | | 0 0 0
P. virgatum 0 0 z 0 | 1 І [01] 0 | 11 1 1 1 0 [0l] 0
Volume 89, Number 4 Aliscioni 523
2002 Filogenia de Paspalum
Apéndice 3. Extendido.
19 20 21 22 23 24 25 26 27 28 29 30 3132 33 34 35 36
Panicum laxum l 0 Z 0 о о [Ol] 2 0 1 0 0 00 0 0 0 0
Eriochloa 0 0 0 0 0 0 1 2 0 1 0 0 0 1 [I] 0 | l
Anthaenantiopsis 1 0 0 1 2 1 [Ol] 1 0 I | 0 0 0 [Ol] [01] [OI] [Ol]
Axonopus 1 0 [Ol] 0 2 1 — 000 0 о 1 о [Ol] [01] 0 0
Thrasya 1 0 [Ol] |! 2 1 — 00 ] 0 0 0 | l l l l
P. malacophyllum l 0 l 0 2 |] — 000 0 0 0 ! l l l l
P. simplex [01] 0 l 0 2 1 — 0 0 0 0 0 0 ! l l | 1
P. bertonit 0 0 1 l 2 1 — 00 ] 0 1 2 0 0 0 0 0
P. lilloi 0 0 l 1 2 1 — 0 0 1 0 1 2 0 0 0 0 0
P. bonplandianum |01] 0 0 0 l 1 — 0 1 1 0 0 1 1 l 1 [Ol]. [01]
P. pallidum 01] 0 0 0 l 1 — 0 10 0 0 1I 1 | | 0 0
P. caespitosum l 0 0 1 l 1 — 0 0 1 0 0 0 0 [OI] [Ol] ! 1
P. indecorum 1 0 | 1 2 1 — 0 0 1 0 0 0 2 l l l 1
P. polyphyllum 1 0 l 0 2 1 — 00 0 0 0-2 2 ] 1 0 0
P. stellatum 1 0 l l 2 ] — 00 1 0 0 1 0 1 1 [01] [01]
P. conjugatum 1 0 | 0 2 1 — 000 0 0 1 O [OI] [Ol] 0 0
P. corcovadense ] о [Ol] 1 [R] 1 0 0 0 I 0 0 0 1 [Ol] 1 [Ol] 1
P. mandiocamum 1 O [Ol] [Ol] [I2] 1 0.0 ] 0 0 0 1 [Ol] ! | |
P. dilatatum 1 0 l l | 1 — 00 I 0 0 1 1 0 [о] ! |
P. urvillei l 0 1 l l 1 — 0 0 1 0 0 ] 1 0 0 | |
P. repens 1 l 0 0 | 1 — 0 0 1 0 0 2 2 1 l l l
P. distichum 0 0 0 l | 1 — 0 0 1 0 0 2 2 0 0 0 0
P. vaginatum 0 0 0 1 | 1 — 0 0 1 о [0] 2 0 0 0 [OL] [01]
P. gardnerianum [Ol] O [Ol] 0 [12] | — 00 1 0 0 1 1 [о [Ol] 1 [01]
Р. ammodes 1 0 0 l | 1 — 00 I 0 0 1l l l l | |
P. erianthum 1 0 [Ol] [01] [I2] 1 — 0 0 I о O 1 0 [OI] [Ol] 1I [OT]
P. equitans 1 0 Jo] 1 [2] 1 — 001 0 0 20 0 0 0 [01]
P. fasciculatum 1 о [Ol] 1 [R] 1 — 0 0 I 0 о 0 0 0 0 1 [Ol]
P. filiforme 1 о — — — 1 — о 0 1 0 0 0 0 0 0 | 1
P. lindnerianum 1 о — — — | — о 0 I 0 о 00 0 0 | 0
P. inaequivalve 1 0 0 0 l 1 0 0 0 1I Q0 0 0 0 [OI] [OI] [01] [01]
P. lineare l 0 — — 1 — 00— — — —0 — 0 — 0
P. proximum 1 0 0 l — 1 — 0 0 — — 1 2 0 0 0 [Ol] 0
P. alcalinum 1 [Ol] ! 1 | 1 — 0 0 1 0 O° I 2 l l l 1
ividum 1 [Ol] | 0 2 1 — 0 0 | 0 о 1 2 l l l l
P. macrophyllum l 0 l 0 2 1 — 0 0 1 0 0 1I 1 0 O [Ol] [Ol]
P. notatum [01] 0 0 1 | 1 — 00 1 0 0 0 0 [Ol] [Ol] 1 0
P. pumilum [01] O [01] 1 |2 1 — 00]! 0 0 0 1 l | [Ol] [01]
P. orbiculatum 0 0 0 0 l 1 — 0 0 1 0 0 1 1 [Ol] [Ol] 0 0
P. juergensii 1 0 0 0 | 1 — 0 0 1 0 0 0 1 [OI] [Ol] 1 |
P. paniculatum [01] 0 0 0 | 1 — 0 0 1 0 0 0 1 [Ol] 0 l l
P. crispatum l 0 1 0 2 1 — 000 0 0 2 2 |] l 1 [01]
P. parviflorum [01] 0 1 1 2 | — 0 0 0 0 0 1 ] 1 ] [Ol] 0
Р. modestum 1 1 0 | l 1 — 0 0 1 0 0 1 1 0 0 0 0
P. plicatulum 1 0 l l 2 1 — 00 ] O O 1 1 [01] [01] | 1
P. coryphaeum [01] 0 1 | 2 1 — о 0 0 0 0 1 1 l l | l
P. quadrifarium 1 0 0 | | 1 — 0 0 I 0 0 1 | 0 0 l l
P. penicillatum 1 0 0 0 | ї — 0 1 0 Q O 1 1 [01] [01] 0 0
P. pygmaeum [01] 0 0 0 l 1 — 010 0 0 1l | l l 0 0
P. saccharoides 0 о [Ol] 1 [l2] | — 0 0 I 0 0 2 0 1 0 l 0
P. regnellii mic 0 6 1 li — 001 9 9 lJT5 NR 1.3
P. virgatum І [Ol] 0 1 l 1 — 0 0 1 [Ol] 0 1 1 [Ol] 0 [I] 0
GOCHNATIA (ASTERACEAE, Susana E. Freire”, Liliana Katinas’, and
MUTISIEAE) AND THE Gisela Sancho?
GOCHNATIA COMPLEX:
TAXONOMIC IMPLICATIONS
FROM MORPHOLOGY!
ABSTRACT
Gochnatia is one of the largest genera of the tribe инен (Asteraceae) and has been traditionally charac ۰
by its homogamous capitula with isomorphic corollas. A morphological study of Gochnatia and associated genera, i.e.,
Actinoseris, Chucoa, Cnicothamnus, Cyclolepis, Hyalis, aa Nouelia, Pleiotaxis, and Wunderlichia, рн
out to evaluate the circumscription of Gochnatia and its sections, and the a of this complex of genera, The
бике studied involve habit, leat баи — ушевсепсе typ Dess тап, and venabionA a 9
5
ST sex, — of oe ee a о es, ache nial pu bee 'ence, end pay а MSN * these
f d: (1) although Gochnatia is highly variable in most of the characters studied. it can be defined by this
suite of a isomorphic to subdimorphic — ulate anther appendages, and smooth style branc hes; (2)
ol
=
>
ы
&
=
=
£z
5
=
РА
=
=
^
D
A
£
=
sections of Gochnatia needed to be re-evaluated. As result of this, two sections, i.e., sect. Discoseris and sect. Penta-
phorus, are confirmed; two е иш аге opel i i.e., sect. Glomerata sect. nov. and sect. Rotundifolia sect. ; three
sections are redefined, i.e., sect. Hedraiophyllum, sect. Leucomeris, and sect. Gochnatia, while sect. Anassnaphiotdes | is
formally published; and (3) the combination of apic — anther appendages and smooth style brane — is unique to
Actinoseris, Cnicothamnus, Cyclolepis, Gochnatia, Hyalis, lanthopappus, and Nouelia within the Mutisieae. This group
of genera is recognized here as the Gochnatia complex, with Gochnatia as the basal genus of this нен Chucoa,
Pleiotaxis, and Wunderlichia do not belong to the Gochnatia complex.
Key words: Asteraceae, Compositae, Gochnatia, Gochnatia complex, infrageneric classification, morphology, Mutis-
ieae.
Gochnatia Kunth is one of the largest genera of 1838; Jervis, 1954; Cabrera, 1971) since the genus
the tribe Mutisieae, subtribe Mutisiinae sensu lato was established by Kunth (1818). Cabrera (1971),
(including Gochnatiinae; Robinson, 1991; Bremer, in his monograph of the genus, divided it into six
1994). It comprises 68 species, nearly all Neotrop- sections: sect. Discoseris, sect. Gochnatia, sect.
ical and 2 endemic to the mountains of southeast- Hedraiophyllum, sect. Leucomeris, sect. Moquinias-
ern Asia. All the species of Gochnatia have been trum, and sect. Pentaphorus. He suggested the ar-
traditionally described as discoid with actinomor- — tificial delimitation of some sections, such as Hed-
phic, deeply 5-lobed corollas, features that are ple- raiophyllum, and in other cases emphasized
siomorphic within Mutisieae (Bremer, 1994). The geographical distribution in distinguishing groups
only apomorphic character suggested for the genus of species, such as in section Gochnatia.
is the acuminate to apiculate apical anther ap- Jeffrey (1967), when relating Gochnatia to other
| I | h 8
»endage of the stamens (Bremer, 1994). This char- taxa, established the Stifftia-series of genera. which
I 8 J 8
acter, however, is shared by other genera of Mutis- all have short, rounded, glabrous style arms, and
іеае. commonly glabrous corollas. The Stifftia-series was
The infrageneric taxonomy of Gochnatia has then divided into four subseries by slight differ-
been much discussed (Lessing, 1832; de Candolle, ences in the shape of the style arms, although these
' We thank Vicki Funk, Per Ola Karis, Richard Keating. Harold Robinson, Peter Stevens, and Tod Stuessy for he Ipful
Ma ин on an earlier draft of this manuscript, and one anonymous reviewer comments. We are espec ially grateful
to John Pr who enthusiastically improved the manuse ript with his suggestions. We are ue — ful to the curators
of G, K, LP. МО, NY, S, Sl, and US for the loan of specimens, and Nadia Roque for access to her prepublication
manuscript on lanthopappus (now published i in Novon). Finally, we thank Victoria Hollowell * 'ause ы editorial w
was extremely helpful, and Víctor H. Calvetti for inking our — pencil illustrations. Support for this study by the
;onsejo Nac ional de Investigaciones Científicas y Técnicas (CONICET), ), Argentina, the Smithsonian Institution, Wash-
ington, D.C. (for G.S.), and the National Geographic Socie ty (grant 5776-96) is gratefully acknowledged.
? División Plantas Vasculares, Museo de La Plata, Paseo del Bosque, 1900 La Plata, Argentina. freire@museo.
fenym.unlp.edu.ar (author for correspondence).
ANN. Missouni Bor. GARD. 89: 524-550. 2002.
Volume 89, Number 4
2002
Freire et al. 525
Gochnatia Morphology
differences were not described for each subseries.
Of these subseries, the Gochnatia-subseries includ-
ed Achnopogon, Cnicothamnus, Gochnatia, Nouelia,
and Oldenburgia.
Cabrera (1971) considered Actinoseris, Cyclole-
pis, and Pleiotaxis closely — to Gochnatia by
their apiculate anthers , he also associated
Chucoa with Gochnatia CREE 19 ansen
(1991) considered only Actinoseris and Cyclolepis of
Cabrera's group to be close to Gochnatia, excluding
Pleiotaxis and Chucoa, and added Hyalis and
Nouelia to what he called the “Gochnatia-group.”
Hansen argued the presence of cone-like involu-
cres, i.e., with light brownish bracts imbricately ar-
ranged and resembling the cone of a spruce as a
probable synapomorphy of the group. Recently,
Roque and Hind (2001) created the new genus /an-
thopappus for the species Actinoseris corymbosa, In
this work, the authors grouped the genera Actinos-
eris, Chucoa, Cnicothamnus, Cyclolepis, Gochnatia,
Hyalis, lanthopappus, and Nouelia by their apicu-
late to acuminate apical anther appendages. Fur-
thermore, Roque (2001) and Roque and Pirani
(2001) re-circumscribed the genus Richterago Kun-
tze according to the genus concept first employed
y Lessing (1830) to include the species of Acti-
noseris and the species of Gochnatia sect. Discos-
—
eris. We consider, however, that the characters used
to distinguish Richterago are widespread in Gochn-
atia or are shared with other genera, such as her-
baceous to subshrubby habit (shared among Gochn-
atia sect. Discoseris, Actinoseris, and other such
related genera as Hyalis and lanthopappus), leaves
rosulate to alternate (alternate leaves are present in
most species of Gochnatia), venation pinnate (pre-
sent in most species of Gochnatia), capitula ho-
mogamous and discoid (present in most species of
Gochnatia) or heterogamous and radiate (present in
some species of Actinoseris), as well as pappus fea-
tures, i.e., pappus uniseriate, with 25 to 42 bristles,
united into a fleshy ring (present in species of
Gochnatia sect. Gochnatia). For these reasons, we
affirm here the traditional concept of Gochnatia and
Actinoseris (excluding A. corymbosa = lanthopap-
pus corymbosus) as was established by Cabrera
(1970, 1971).
Some phylogenetic studies relate Gochnatia with
other genera of Mutisieae; for example, Jansen and
Palmer (1987) related Gochnatia with Ainsliaea,
Onoseris, and Stifftia, and Karis et al. (2001) relat-
ed it with Mutisia and Trixis. However, because of
the few taxa sampled in the tribe, these analyses
are not considered here to address relationships at
the generic level.
In other analyses (Karis et al., 1992; Jansen &
Kim, 1996), Gochnatia was found to be an isolated
taxon within Cichorioideae. The cladogram of Karis
et al. (1992), based on morphological characters,
shows Gochnatia as sister to most Asteraceae, ex-
cluding Barnadesioideae and four genera of Gochn-
atiinae. The ndhF tree of Jansen and Kim (1996)
also shows Gochnatia as an independent lineage
positioned basal to most Asteraceae, excluding Bar-
nadesioideae and the core of Mutisieae examined.
ere are other genera within Mutisieae such as
Quelchia and Neblinaea from the Guayana High-
land with apiculate anther appendages, but mor-
phological features of these genera (Pruski, 1991;
Bremer, 1994) mark a departure from Gochnatia
and associated taxa. On the other hand, the plan-
altive Brazilian genus Wunderlichia, with shortly
apiculate anther appendages, was either placed in
(Pruski, 1991) or it was
considered isolated within Mutisieae (Hansen,
1991: 1992; Bremer, 1994).
The group selected here for the analysis is main-
the “Stenopadus group”
Karis et al.,
ly represented by the genera with apiculate anther
appendages apes, by Jeffrey (in part, 1967),
Cabrera (1971, 1977), Hansen (1991), and Roque
and Hind al It is comprised of Actinoseris sen-
su Cabrera, excluding A. corymbosa (7 species),
Chucoa (1 species), Cnicothamnus (2 species), Cy-
clolepis (1 species), Gochnatia (08 species), Hyalis
(2 species), lanthopappus (1 species), Nouelia (1
species), Pleiotaxis (ca. 25 species), and the con-
troversial Wunderlichia (6 species).
There are several potential diagnostic features in
Gochnatia and relatives such as the habit, leaf mor-
phology, types of capitulescence, involucre, florets
trichomes, and pappus that have never been stud-
ied comparatively or in detail. A detailed morpho-
logical study of Gochnatia and its associated taxa
is needed as a first step to provide a robust base
for discussion.
On the basis of morphological evidence, the
goals of this study are: (1) to evaluate the circum-
scription of Gochnatia and its sections, and (2) to
eroup Gochnatia with other genera based on their
similarities.
MATERIAL AND METHODS
Herbarium specimens of the studied taxa, i.e.,
Actinoseris, Chucoa, Cnicothamnus, Cyclolepis,
Gochnatia (64 of its 68 species), Hyalis, lantho-
pappus, Nouelia, Pleiotaxis, and Wunderlichia (Ap-
pendix 1), were examined to assess characters used
previously to distinguish taxa in this group of gen-
era, and to search for additional characters. The
data were augmented by literature studies (Fran-
526
Annals of the
Missouri Botanical Garden
chet, 1888; Cabrera, 1950, 1951, 1955, 1970,
1971; Jeffrey, 1967; Barroso & Maguire, 1973; Zar-
dini, 1975; Sancho, 1997, 2000; 1997;
Roque & Hind, 2001; Roque & Pirani, 1997,
[
Roque,
Vegetative and floral parts were dissected and
observed after boiling in water and stained with 2%
safranin. Freehand sections of leaves were per-
formed and these sections examined to determine
the presence of hypodermis, and then stained with
safranin. Drawings were made by the authors using
E
stereomicroscope Wild M5 and
ё microscope
Leitz SM Lux with the camera lucida technique.
Apical anther appendages were described and
drawn to include the portion ranging from the the-
cae apex to the apex of the stamen.
Descriptive terminology for the trichomes follows
Ramayya (1962),
synonyms of
Whenever possible, additional
trichome terminology were added
(Payne, 1978; Font Quer, 1979; Moreno, 1984;
Harris & Harris, 1994; Metcalfe & Chalk, 1950).
Some authors use the term “trichome,” whereas
others apply the term “hair” in their pubescence
classifications. Both terms are cited here following
each author classification.
RESULTS
For
Gochnatia with their corresponding species accord-
ing to Cabrera (1971) are listed in Table 1.
comparison with our results, sections of
VEGETATIVE CHARACTERS
Habit. Таха under study display four types of
habit: small trees, shrubs, subshrubs, and perennial
herbs. Several species of Gochnatia (e.g.. G. arbo-
rescens, G. decora, G. elliptica, G. ilicifolia, G. mag-
na, G. oligocephala, G. palosanto, G. polymorpha,
G. spectabilis, G.
Wunderlichia
tortuensis), Cnicothamnus,
Most
and
and
are small trees. species of
Nouelia are
shrubs. Gochnatia sect. Discoseris, Hyalis, and lan-
Gochnatia, Chucoa, Cyclolepis,
thopappus are subshrubs. Actinoseris and Pleiotaxis
are perennial herbs. Cyclolepis is the only spinose
genus and is aphyllous at maturity.
Leaves. All taxa analyzed have alternate, oc-
casionally rosulate, subsessile to shortly petiolate
leaves.
Leaf consistency: The leaves are coriaceous or
subcoriaceous, but only some species have leaves
with an adaxial hypodermis. It is absent in some
species of Gochnatia (e.g., G. amplexifolia, G. are-
quipensis, G. cardenasii, G. discoidea, G. foliolosa,
G. glutinosa, G. hypoleuca, G. intertexta, G. magna,
G. palosanto, G. purpusii, G. rotundifolia, G. ver-
nonioides), Actinoseris, Cnicothamnus, Cyclolepis,
yalis, lanthopappus, and in some species of Pleio-
taxis (e.g.. P. eximia, P. huillensis). Occasionally,
the hypodermis is discontinuous (e.g., Gochnatia
argentina, G. discolor, G. orbiculata, G. ramboi).
Leaf shape: The leaves are always simple ir
the taxa under study, but show great variation in
shape, from linear to suborbicular. Most species of
Gochnatia sect. Gochnatia and Nouelia have leaves
ovate to ovate-elliptic, obtuse to subobtuse at the
apex; occasionally ovate leaves are acute at the
apex (e.g.. G. polymorpha, G. vernonioides), or cor-
date at the base (e.g., G. cordata, G. haumaniana)
(Fig. 1A). Most of the Caribbean species of Gochn-
atia sect. Gochnatia, Actinoseris (A. hatschbachii, A.
polymorpha, A. radiata), Chucoa, lanthopappus,
and Wunderlichia have obovate leaves (Fig. 1B).
Elliptic leaves (Fig. 1C) are rounded in the base
and apex in G. discoidea, attenuate at the base and
apex in G. magna, G. sect. Leucomeris, Actinoseris
(A. polyphylla), Cnicothamnus, and some species of
Pleiotaxis (e.g., P. huillensis, P. newtonii, P. selina,
P. subscaposa). Gochnatia sect. Pentaphorus, G. ar-
gyrea, Actinoseris (A. angustifolia, A. stenophylla),
Cyclolepis, Hyalis, and most species of Pleiotaxis
(e.g.. P. ambigua, P. decipiens, P. dewevrei, P. li-
nearifolia, P. rogersii) have linear leaves (Fig. 1D).
Leaf margin: Only the Caribbean species of
Gochnatia sect. Gochnatia (except G. crassifolia
and G. picardae) and Chucoa have spinose margins
(Fig. 1E). In the remaining taxa the leaf margin is
predominantly entire or denticulate (Fig. 1F).
sal venation: The venation is predominantly
pinnate, although in taxa such as Gochnatia are-
quipensis, G. glutinosa, G. rotundifolia, Hyalis, and
lanthopappus the leaves are three-veined (actinod-
romous).
CAPITULESCENCE
Capitula are borne singly or 2 or 3 at the end of
the branches or may be clustered in pseudoco-
rymbs, pseudoracemes, or pseudopanicles, in open
to condensed or glomerulose synflorescences de-
pending upon the length of the peduncles.
Solitary or few capitula (2 or 3) are short-pedun-
culate to glomerulose in Gochnatia sect. Gochnatia,
Cnicothamnus, Nouelia, and some species of Wun-
derlichia (W. azulensis, W. crulsiana, W. mirabilis)
(Fig. 2A). They are scapiform, long-pedunculate in
Gochnatia suffrutescens, some species of Actinoseris
e.g., A. angustifolia, A. hatschbachii, A. polymor-
pha, A. radiata),
—
Chucoa, and some species of
Pleiotaxis (e.g.. P. subscaposa) (Fig. 2B).
527
Freire et al.
Volume 89, Number 4
2002
Gochnatia Morphology
(q 2d4) y ada) sndded :sausyor ur sarey
їрүприрүз ayeyideo ((әүетштәв]) spre; оош» “(әле
-noide Apdniqe) apne? o3epuadde ләцир :рәдо|
-c Ájdəəp Аләл о} Аүдәәр se[[o100 :(e 10 Z) e]
-mudes Алгүов “шлеш (әлциә) asourds YIM soAg»]
g edá1 sndded ‘рез 9jeruroe[ *91enuo]
-e 93epuodde ләццир :a]« se[[o100 :surg eur au
үр aje[orpro “SOLIR[[AYd snosqepsqns YIM әләпүол
-u1 te[njrdeo Але ISNOIGR]S "рәшәл-әәлу SAARI]
g 2d4 sndded :s[1e] ajeruroe[ “(aypenoide
Арапе) ayepneo osepuedde toque (е 40 Z) e]
-njdeo Are [Os :рәшәл (-9291g1qns) Ayeyeuuid SOAR’
y edaq sndded :s[ıe ature] *aje[noide Apdnaqe
a3epuadde tsqui&i000pnosd asodeos ш sno1ouinu
10 *oje[nounpod-zuo| “Á1B]1]OS e¡mideo “SQnIysqng
UB9q4 UB”)
eqno)
ә паә URJIUTWO
iqndoy ueorunuoq
eqn’)
ueaqq.ie")
eqn’)
rag s
nd N
eunuoziy MN mjog S
ernog
enog
рцорәлр “x
т1тәшоЎ 74)
Di]upoauua °“)
21111112 “)
nunuyə °“)
sisuaqno °“)
D1 jofissp49 4)
sapionjdpajspuy “1998 DIJDUYIO?)
DIJOJIPUNIOL 2) y,
Dijo fipunioy “1998 DIDUYIOS)
sapio1uouA424 ©)
пкр °<)
puizpjnd °<)
n4opfiaamo °<)
NSDUIPADI °<)
DUDIAO %9
sisuadimbaap “9
DIDUYIOY *1238 DIIDUYIOY
sua2sajnaffns °<)
DIPIOISIP °“)
piofixojduim 1)
S142980081(] 295 DDU yoo)
“вә” 0170/1211 `2)
—
шегү Y 5їллә[ (uoo) 11220102. ‘<
шегу (oxe[g Y `$) pyrunauua `t
ure y (идә) noudijo °“
шерү Y swf qq) nuvwyə °<
шерү Y Swf (erquae7)) sisuaqno °“)
шүү
y suf (чори) DIJOfISSDALI °<)
UB] Y Y smf (uopug) 77720102. °<)
шегу Y smf (UOWLIG) 270212703 °<)
zougui[( (191) nyong 4
шерү Y smf (пори) DIDNUIND °<)
_
:sə19əds ивәфивә
‘5597 pijofipunjoa 9
yyuny уәр1о1иоиләа x
axe[g `{ `$ MSDUIPIDI °<)
Seg A ND д
"“mpues sisuadinbasp “9
:sər9əds ивоәшү YINOG
Di]Dui204) “WIS DIJDUYIOS)
8.19.1Q8”) susyso mans 1)
n
~
юләлЧЕТ)
e1o1qe7) (1әирлее)) ю1уоїхә]4шъ °<)
A
‘SSI J) DIPIOISTP “) x.
STLISOIST(] 029% DIJDUYIOL)
вләәвлецә 21]souser(]
UOHNLYSI(]
Apnis sių} 0} BUIPIOIIE suonoag
(1261) &191qe7) nsuas SUOT}IIG
'uorpas YORA JO sotveds adA] = y ‘sosoyjuoird ш 3uneodde
so1njeoj пошшоәип YIM “paptacid әле uonoos YORE 10] SISJORIBYO onsouzerp pue uonnquysip olydeldos93 әчү “(0007 “OYouRs) pjpidaomsuap “) pue (pr6[ Paqe) торур
4) jo uonrppe эці чим "([76[) &1o1q97) Aq рәишцәр suoroes əy) чим poreduioo *A3oJoudaour əy} Jo siseq əy} uo &pnis siy} ur pausi[qeiso. 927041902) шцим ѕиоцоә `| әре,
Annals of the
528
Botanical Garden
issouri
M
g 2d4 snd
-ded :(y100us) sje} ayeruroey *(ayepnorde Apidniqp)
ayenuayye a2epuadde Joyque :(морәќ) INYM
(ZL 01 8)9 01 y s1219g :o[ounpad əy} our Surpua]
-хә ѕәпе|үАҷа ш#теш әці 1? IPJO *snoaqe[3
Чим әлоп[олш :saporuedopnasd 10 вәшәәвлорпәва
‘squAıo20pnasd әѕо[пләшо[8 ut epnyideo sno1auny
OOIX2[N
PIAI[Oq S “eunuadly N
OSA UWS S
(y adá1) Я addy snd оотхәр
-ded :S[18] щоошѕ (әјепиәце) a1e[noıde Apidniqe
asepuadde sayjue :(0)с)0) o1 zT S1310 :squi&i
-ooopnasd aso[nıau0[3 ut 1911920] мә} e[niide7)
OOIX2]N
ODIXIA
ивэддив7
eqn)
eqn)
тш 7)
отирѕотра `£)
ponajodty °<)
SILIWOINA] "29s DDU 1204)
(u0noas Jayjouer 0] рәлош)
SUIISILOQLD су,
D]D42UI0]4) "1998 DIPDUYIOS)
nuospn e)
SISUINILO] °4)
чәјоцѕ 9
DUDIDIGDS °<)
DANIIL 74)
әррірла °<)
Dsojnosogfionvd ^)
торага “4
DYJUDIAO °<)
pijofisnigo “5
pupjuow `£)
оуоцә2о1лш ғ)
sisuanjupui
р]хә]лә]и1 °<)
'шиәәлсу X UOSUIGOY nynus °<)
e1a1qe7) озирѕотра `)
ABI “Y (90) Donajodty “9
әәдәриғія "< `], usndind «)
?IoIqe7^) puspu 4)
`55ә^] nmpa4o2 °“),
әәЗәрирд `$ ||, suaosa40quD 4)
штї]]ХцЧолюлрәр 71998 DIDUYI0y
UIRTY Y smf (uogyug) nuospan е)
zouguif (917) sisuanz40] ©)
шегү Y smf (uonug) Lafpys “4
UIR[Y Y stAJ9(. DUDIDLIDS °<)
uey Y smf (uonug) 011221 °<)
ZaugUIf (q1) eppavoid “9
£131q*7) ('2u91H) Dsoynosopfionnd ©)
ute[y Y smf (uonug) pijofiaund e)
pıemoH (91) DYJUDFII0 °<)
шү
Y smf (uonug) vyofisniqo °4
шегу Y smf (uonug) DUDJUOU *)
IN
y swf ("qasuz)) рурудәәоләтш =)
шү
Y Swf ("qosue)) sisuenjupw °<)
uey Y тллә[ (uoo) риртѕтош 7)
UIRTY Y smf ("әѕце)) юухәмәйи “9
вләўәвлецә onsouzeT uonnquisi]
Ápnis sip 0] Зитрлодәр suonoag
(1261) &191qe7) nsuas suonoag
'penuguo) `I [qe]
529
Freire et al.
Volume 89, Number 4
2002--
Gochnatia Morphology
q 2d4 sndded [тв] әјвтшовү *Цзоошѕ
“ae pnoide Apidniqe a3epuadde ıaqıue :әвү
10 IHYM “(OZ 01 QT 10 *) *€) € SIALOY SIWI
-opnasd 4jeo[ “әѕојпләшо[8 ur e[njtdeo ѕполәшпи
euquagiy M
9114, [Pus
:әјеүприве[8 *paura4-221gj *Ap[9jeuutd *saAea[ 1eaurq
[Ze1g S
peg 5
PIAIT[Og ‘nag Jo sapuy
pzeig S
Aengeıeq ‘zeg ©
eunuesiy *Aenzni(] ‘Aenseieg “лея
[Ze AS
[zeig 4
[2219 Y
[Zeg S
|izeag "AenSe1eq
[12219
[12819
7) od) sndded :s[rej оош» *9]RIULOR] °21] pzeig
-noide храпе ozepuadde 1aque :(orqdaouiost) pzeig
orgdaouipqns se[[o100 :(proosrp) unoprosrp ‘(snow pzeig 4l
-рдошоц) snowue301919y e[njrdgo :so[oruedopnosd eunuoziy *Aen2ni(] ‘рар *^enzeieq
3500] 0} әѕо[пләшо[8 ut e[nirdeo вполәшпи ‘siey pzeig я
(әтеәзеру) рәшле-с 01 -e “g YIM вәлвә{ :(snoto Aengeıeq ‘zeg S
-әоцош) snowRsAjod *snoroaorpou&2 “піце 10 S221], [zeig S
Purjuo21y N
esy
EISV
Dsou11n]8 °<)
psojonof `9
snioydnjuad “1998 DIIDUYIOS)
ри1тјәа °<)
pydiowsjod ‘4)
Dip navund ©)
D]p]n21Q40 °<)
pnjoudoooZi]o 4)
punssijjow °<)
DUDIUDUNDY °<)
B191GB") 1Y9DQYISIDY 74)
1L9UPADI °<)
ppunquojf 1)
4OJOISIP E
OYOURY 4) (8191427) pypydasisuap 19
5597] DIDPLO) 9F
DUDIJIYIUDIG `0
DUNUITID °<)
штп]]Хцдотюлрәц 152s DIDUYIO)
sinqpi2ads 4)
D4029p ©)
шү Y XooH (uoq `@) nsouimj2
2
шү Y оон (чод O) 80701/0f 9x
snioydpjuag "jes D11DuY4204)
e1a1qe7) ("uog) nu1ngoa
R.191QB (sso) Dp1pa4os *
P191QB7) DUDAGSNL
PIOIqP^) 10QUUD4
e1o1qe7) vayopnd
e191q87) ('sso']) DydiowAjod
e1a1qe-) ("ә") рурупәлита *
RIOIGe’) (әшүрү) 220101410 °
&1a1qe7) (12upae4)) DIDYdav037]0
E191{E7) (эщ[ер) Dumssioui
EI23JQE7) DUDIUDUTIDY 2
2.1918 (128g) тзәира›З
РЛӘЛЧЕТ) эри nquoy *
ЛӘХРЧ 10]02%р
вләлв”) (9q) Pupljeyoun]q *
BIOIGe’) (QUEJA) рәгс *
еләге”) (E1a1qe7)) puijuazan °
p,
UO
So
y
SESS
umajspiuimmbopy "2998 DI]DUYIO*)
sso] (uoq `(]) 81729072905 2) y,
e.191G8") (Zin M) D4022p
s149u102Tio'] 90295 DIJDUYIOY)
9
вләәвтецә oysousei(y попи]
Apnis stu 0} urp1oooe suonoag
(1261) 191087) nsuas suonoag
penunuo) "р [QPL
530 Annals of the
Missouri Botanical Garden
0
Pro Pde Cge
Gfo Gar Hla Aan
i Gpo Gro С
Figure 1. Leaf shape. —A. Ovate (le fti to rigai к» d rotundifolia (Handro 157, LP), G. ramboi — 50005
S), G. КОО (López & Sagdstegui 3354, LP), G. cordata (Serrano 6, LP), Nouelia insignis (l'Abbé D 2498,
US). —B. Obovate (left to right): Wunderlic e crulsiana — et al. 2615, MO), Gochnatia intertexta A-1680,
Ag G. pauciflosculosa (Wilson 7428, К), G. seg cae Ekman 4023, S), lanthopappu e s & Cuezzo
2304, LP), Actinoseris radiata (Hatsc hbach 690, LP), Chucoa ilicifolia (López Miran P). — “Elliptic (left to
— Gochnatia discoidea (Blanchet 3345, LP), G. magna (Cronquist 11277, N
ү), G. dites e е) 32, LP),
Volume 89, Number 4
2002
Freire et al.
Gochnatia Morphology
Few to several capitula (more than 4) arranged
in short-pedunculate to glomerulose pseudoco-
rymbs are found in Gochnatia arborescens, G. mag-
na, G. purpusii, Gochnatia sect. Leucomeris, Hyalis
lancifolia, and Wunderlichia (W. bahiensis, W. insig-
nis) (Fig. 2C). Numerous capitula arranged in scap-
iform, long-pedunculate pseudocorymbs are found
in Gochnatia sect. Discoseris (except G. suffrutes-
cens), Actinoseris (А. stenophylla), Hyalis argentea,
lanthopappus, and Wunderlichia senaeii (Fig. 2D).
sochnatia sect. Pentaphorus and Cyclolepis have
numerous capitula borne in short-pedunculate to
glomerulose clusters, that are in turn arranged in
leafy pseudoracemes (Fig. 2E).
compact or glomerulose clusters, which lack leaves.
at the tip of the branches, are characteristic. of
Gochnatia palosanto (Fig. 2F). Loose pseudora-
cemes are present only in some species of Pleio-
laxis (e.g.. P. angusterugosa, P. gombensis, P. oxy-
lepis, P. racemosa) (Fig. :
Numerous
Pseudoracemes in
-
capitula — in glomerulose
pseudopanicles are present in most species of
Moquiniastrum, in G. cordata, G.
hypoleuca, and G. smithii (Fig. 2H). Short-pedun-
culate (Fig. 21) or loose pseudopanicles (Fig. 2J)
are present in species of Gochnatia sect. Moqui-
Gochnatia sect.
niastrum (G. argentina, G. argyrea, respectively).
INVOLUCRE
Shape. The involucre shape is either oblong to
campanulate or turbinate. Oblong to campanulate
of most species o
Gochnatia (Fig. 3A), Chucoa (Fig. 3B), Cnicoth-
amnus (Fig. 3C), Cyclolepis (Fig. 3D), Hyalis (Fig.
involucres are characteristic
=
3E), Nouelia (Fig. 3F), Pleiotaxis (Fig. 3G), and
Vunderlichia (Fig. 3H). Turbinate involucres, on
the other hand, occur in a few Caribbean species
of Gochnatia sect. Gochnatia (e.g., G. cubensis, G.
intertexta, G. pauciflosculosa), in some species of G.
sect. cee died (G. hypoleuca, G. palosanto,
G. smithii) ( 3D, G. suffrutescens,
and /anthopappus (Fig. 3J).
Fig. Actinoseris,
Size. The involucre size ranges from 2 to 45
(2—)4—7(—8)
mm high, with three main categories:
mm, 10—18 mm, and 20-45 mm high.
(A) The smallest involucres [(2-)4-7(-8) mm]
are present in Gochnatia sect. Pentaphorus, Gochn-
alia sect. Moquiniastrum (except G. argyrea with
9-10 mm), G. hypoleuca, G. microcephala, G. pa-
losanto, G. smithii, Cyclolepis, and Hyalis.
(B) Intermediate involucres (10-18 mm high
are present in some species of Gochnatia (e.g.. G.
arborescens, G. cordata, G. discoidea, G. ilicifolia.
G. magna, G. patazina, G. purpusu, G. recurva, б.
rotundifolia, G. vernonioides),
М
Actinoserts, Chucoa,
and /anthopappus.
(С) The biggest involucres (20—45 mm) are dis-
played by a few Caribbean species of Gochnatia
Gochnatia (e.g., G. G. G.
ekmanii, G. picardae, G. sagraeana), Cnicothamnus,
Nouelia, Pleiotaxis, and Wunderlichia.
sect. cowellii, cubensis,
Series of phyllaries. The phyllaries are arranged
in several imbricate series. The highest number is
(6)7 to 10 series, and this is found in a few Carib-
bean species of Gochnatia sect. Gochnatia (e.g.. б.
cowellii, G. cubensis, G. picardae, G. recurva), б.
arborescens, Cnicothamnus (Fig. 3C), Nouelia (Fig.
3F). Pleiotaxis (Fig. 3G), and Wunderlichia (Fig.
3H). Three- to six-seriate involucres occur in most
species of Gochnatia (Fig. ЗА), and Actinoserts,
Chucoa (Fig. 3B), Cyclolepis (Fig. 3D), Hyalis (Fig.
3E) and /anthopappus (Fig. 3J). Gochnatia sect.
Leucomeris and three species of G. sect. Hedrato-
phyllum (G. hypoleuca, G. palosanto. G. smithii)
show a peculiar feature in their involucres. The ca-
pitula are 3- to 4-seriate, but the peduncle bracts
extend into the involucre giving a 7- to 12-seriate
condition (Fig. 31)
Phyllary pubescence and shape. In almost all
the taxa studied the phyllaries are dorsally pubes-
cent or subglabrous. However, Gochnatia sect. Leu-
comeris, G. hypoleuca, G. palosanto, G. smithii, and
G. rotundifolia have dorsally glabrous phyllaries
with ciliate margins.
ost genera have linear to oblong or obovate
phyllaries with entire margins. Cnicothamnus has
phyllaries with an apical appendage rounded
rhombic, with lacerate margins (Fig. 3C), and Wun-
derlichia has phyllaries with scarious, colored, and
occasionally fimbriate margins (Fig. 3
€
Actinoseris polyphylla (Hatschbach 35304, LP), Cnicothamnus lorentzii А Жет "
10780, US). —D. Linear (left to right): Gochnatia foliolosa (Cabrera 34
Hyalis lancifolia (Sc hinini — LP), Actinoseris angustifolia ( (Hatsc m 1 — Б A Pleiotaxis rogersii (Re
US (Rojas 7104
;ochnatia montana (Ekman 18725,
),
5), P. dewevrei — S), Cyclolepis genistoides
E
Chucoa ilicifolia (López yr 1090, e
S). G. cubensis (Ekman 9632, S), G. intertexta (Alain A-
Pereira 8609 & Pabst 7984, 1, р), G. rotundifolia | (Handro 15
—
Р ). Pletotaxis huillensis (Gossweiler
argyrea (Hatschbach 9578, LP).
Чуп 1568,
P). E margin. —E. Spinose (left to Pon
‚ G. enneantha (Ekman Н-1549
^ Dentice — to entire (left to right): G. —
LP), G. foliolosa (Cabrera 3451, LP).
LP )
532 Annals of the
Missouri Botanical Garden
С D
Figure 2. Types of capitulescence. Solitary or few capitula short-pedunculate. —B. Solitary or few capitula long-
8 YF у I 8
к culate and scapiform. —C. Р — short-pedunculate. —D. Pseudocorymbs — -pedunculate and scapiform.
Pseudoracemes glomerulose and leafy. —F. Pseudoracemes apically glomerulose. —G. Loose pseudoracemes. —H.
Ps "edd xinicles glomerulose. —I. Ps E anicles short-pedunculate. —J. Loose pseudopanicles.
F B I | | I
PALEAE (1991) included it in the “Stenopadus group.’
Cabrera (1971) described Gochnatia as typically
Wunderlichia is the only genus in the group — epaleate, and only exceptionally with some pa-
with receptacular paleae. For this reason, Pruski leae, but mentioned no species. However, we
Volume 89, Number 4
2002
Freire et al.
Gochnatia Morphology
10 mm.
40 mm.
Figure 3. Involucre. A-H. Oblong to campanulate. —A.
|
Chucoa ilicifolia (López Miranda 1090, LP). —C. €
(Tinto 2038, LP). —E. Hyalis lancifolia (Schinini 1
rogersii (Rolyns 1568, US
palosanto (Schreiter in 1925, LP
—
— la
found no evidence of paleae in any species of
Gochnatia examined.
FLORETS
Corolla
ment.
morphology | and sex arrange-
Corollas in Gochnatia and allies may be
isomorphic, subdimorphic, or clearly dimorphic.
10 mm.
inicothamnus lorentzii (Padaci 84, LP)
6098, LP). —F. Nouelia insignis (Maire 2516, NY). —G. Pleiotaxis
Wunderlichia crulsiana (Ratter et al. 2615, MO).
‚ —J. lanthopappus corymbosus (Palacios & Cuezzo 2304, LP)
10 mm.
Gochnatia floribunda (Hatschbach 43151, NY). —B.
. —D. Cyclolepis genistoides
І, J. Turbinate. —I. Gochnatia
LP).
All florets of the capitula may be hermaphroditic
or female (homogamous capitula), or the capitula
may have central hermaphroditic florets and mar-
ginal female florets (heterogamous capitula). These
features are combined as follows:
(A) Corollas isomorphic. Plants with discoid and
homogamous capitula made up of hermaphroditic
florets with isomorphic tubular corollas are char-
533
534
Annals of the
Missouri Botanical Garden
acteristic of most species of Gochnatia (Fig. 4A, B),
Chucoa (Fig. АС), Pleiotaxis (Fig. 4D), and Wun-
derlichia (Fig. 4E). Only б. argyrea of section Mo-
quiniastrum has isomorphic tubulose corollas and
heterogamous capitula (Sancho, 2000). Very occa-
sionally marginal female florets (with staminodes)
are found in Chucoa (Cabrera, 1955, and our own
observations).
(B) Corollas subdimorphic. Such corollas are
present only in Gochnatia sect. Moquiniastrum (ex-
cept G. argyrea and G. gardneri) (Fig. 4F) and Cy-
clolepis (Fig. 4G). Gochnatia sect. Moquiniastrum
includes monoecious, polygamous, and gynodioe-
cious plants (Sancho, 2000). Polygamous and gyn-
odioecious species have disciform capitula made
up of subdimorphic corollas, i.e., central hermaph-
roditic florets with tubular actinomorphic corollas
and coiled lobes (Fig. 4F,). and outer, female florets
with tubular subzygomorphic corollas and erect or
coiled lobes of different lengths (Fig. 4F,). In the
gynodioecious genus Cyclolepis the female florets
have tubular-filiform corollas (Fig. 4G,), and the
hermaphroditic florets have tubular corollas (Fig.
4G,).
(C) Corollas dimorphic. Plants with radiate, ho-
mogamous capitula comprising dimorphic, her-
maphroditic florets, with outer bilabiate or subbi-
labiate and central tubular corollas, are present in
some species of Actinoseris (e.g., A. polyphylla, A.
revoluta, A. stenophylla) (Fig. 4H), Cnicothamnus
(Fig. 41), Hyalis (Fig. 4J).
Plants with heterogamous capitula and the type of
—
—
and Nouelia (Fig. 41,
dimorphic corollas previously mentioned are found
in lanthopappus (Fig. АК) and some species of Ac-
g., А.
diata); the outer florets are female and the central
tinoseris (e. arenaria, A. hatschbachii, A. ra-
florets hermaphroditic. These truly dimorphic co-
rollas are absent in Gochnatia.
It may be noted that tubular, deeply lobed co-
rollas with coiled lobes (reaching up to 2/3 of the
corolla) are present in most species of Gochnatia
(Fig. 4A), and in Actinoseris, Chucoa (Fig. 4C), Cni-
cothamnus, Cyclolepis, Hyalis, lanthopappus, Noue-
lia, and Pleiotaxis (Fig. 4D). Tubular, very deeply
lobed corollas with straight lobes (more than % the
length of the corolla, almost reaching the base) are
present in most Caribbean species of Gochnatia
sect. Gochnatia (e.g., б. calcicola, С. cubensis, С.
gomezii, G. oligantha) (Fig. АВ). Bilabiate corollas
have an external 3-dentate lip and an internal 2-
Actinoseris revo-
Nouelia) (Fig. 4J—L).
and subbilabiate corollas have an external 3-den-
4, Acti-
cleft lip (3+2 arrangement, e.g.,
luta, Hyalis, lanthopappus,
tate and one entire, internal lip (3+1, e.g
noseris hatschbachit,
(Fig. 4H,
А. polyphylla, Cnicothamnus)
Number of florets per capitulum. The number of
florets per capitulum ranges from 4 to 300. A
species of Gochnatia (G. crassifolia, G. magna, G.
recurva, G. rotundifolia, G. sagraeana) and most
species of Actinoseris, Cnicothamnus, and Wunder-
lichia have 50 to 300 florets per capitulum. Some
species of Gochnatia (e.g., G. amplexifolia, G. bo-
liviana, G. cordata, G. cowellii, G. patazina, G. ver-
nonioides), some species of Actinoseris (e.g., A.
hatschbachii, A. revoluta), lanthopappus, and Noue-
lia have capitula with 30 to 50 florets. Some spe-
cies of Gochnatia (e.g., G. arequipensis, G. carden-
ilicifolia, G. purpusii), G. sect.
Moquiniastrum, Chucoa, and Cyclolepis have 7 7 to
asi, б.
30 florets per capitulum, whereas some species of
Gochnatia (e.g., G. hypoleuca, G. smithii), G. sects.
Leucomeris and Pentaphorus, a few Caribbean spe-
cies of G. sect. Gochnatia (e.g., G. calcicola, G. cub-
ensis, G. gomezii, G. oligantha), and Hyalis have
capitula with 4 to 6 florets.
Corolla color. The corollas in Gochnatia are
predominantly yellow or cream. In some species
(e.g.. G. rotundifolia, G. decora, G. spectabilis, G.
foliolosa) they are white and/or lilac. In others (e.g.,
G. cowellii, G. ilicifolia, G. intertexta, С. mantuen-
sis, G. sagraeana) they are orange. The corollas are
white, pink, or purple in Actinoseris; in Cnicoth-
amnus they are orange; in Chucoa and Cyclolepis,
yellow; in Hyalis lilac;
taxis deep red, pink, white, or cream; in Wunder-
in Nouelia white; in Pleio-
lichia white or yellow. lanthopappus shows white
marginal corollas and purple disc corollas.
ANTHERS
As well as most species of the tribe Mutisieae,
all the taxa in this study have an apical anther
appendage and basal tails. The apical appendage
can be viewed as an adaptation to protect the pollen
in the anther tube from moisture and insect pred-
ators until the stigma and style push upward
through it for pollen presentation (Stuessy et al.,
1996).
With the exception of Chucoa, with acute anther
appendages (Fig. 5A), anther appendages in Gochn-
atia and related taxa have usually been described
as apiculate. Pleiotaxis differs in its thickened and
bulbous apical appendage (Fig. 5B).
Analysis of this feature in all the genera under
study revealed further variation in its shape that is
particularly useful in grouping species within
Gochnatia. Vhe appendage can be short (anthers
Volume 89, Number 4 Freire et al. 535
2 Gochnatia Morphology
Figure 4. A-E. Isomorphic corolla. —A. Gochnatia — (C — 11277, NY). —B. G. e — ulosa (Eggers
3866, К). —C. Chucoa ilicifolia — Miranda 1090, LP). —D. Pletotaxis diu evrei " 21 —E. Wun-
derlichia mirabilis — еШ & Stutts 999, NY). F, e Subdimorphic corolla. —F. Gochnatia оу А (Hashimoto
624, LP): F,, disc corolla, F,, marginal — —G. Cyclolepis genistoides (Cabrera 3782, LP): G,, marginal corolla.
G,, dise corolla. H-L. Dimorphic corolla. —H. ринин polyphylla (Hatschbach 35304, LP): H,, disc corolla, H,,
marginal corolla. —I. Сп e lorentzii (Cabrera et al. 22576, LP): I, marginal corolla, L,, disc corolla. —J.
Hyalis lancifolia (Schinini 16098, LP): J,, marginal corolla. J;. disc corolla. К. —— corymbosus —
Cuezzo 2304, LP): K,, marginal m K,, disc corolla. —L. Nouelia insignis (Maire 2516, NY): L,. marginal coro
L,, disc corolla.
536 Annals of the
Missouri-Botanical-Garden -
Gpar
Gpat
Wi A
Sect. —
Clo
| ire Cge
AN Gde Gsp
Gro
E F
is ou}
PA AAA NA
Sect. Discoseris Sect. TT Nin Waz
Gar
ES Sgl Gco
Ja
Sect. Pentaphorus
0,3 mm.
K L
Figure 5. Apical anther appendage. —A. Not apiculate: — ilicifolia (López Miranda 1090, LP). —B. Apiculate
bulbous: Pleiotaxis eximia (Rolyns 1836, US). C-F. Caudate anther appendage. —C. Gochnatia sect. Сос "o —
Cabre ra left to right): Goc — fp nasii (Cordo & Ferrer 88-B-17, SI), G. cowellii (Britton & Cowell 1
urviflora (Jeréz et al. 49120, LP), G. ilicifolia (Small & — 8526, К), С. montana (Ekman 18725, *
nant (Shafer 2938, NY), С. e s & Sagástegui 3409, LP), G. recurva (León 20946, LP). —D. Actinoseris
T
Volume 89, Number 4
2002
Freire et al. |... . .
Gochnatia Morphology
apiculate) to very long (anthers apically caudate).
In addition, it can abruptly terminate in a sharp
point, or gradually taper above into an attenuate
point. The following three combinations were ob-
served:
(1) Apically caudate (Fig. 5C-E): in most spe-
cies of Gochnatia sect. Gochnatia (except G.
boliviana, G. buchii, G. microcephala, G. tortuensis,
and б. vernonioides with abruptly apiculate apices)
(Fig. 5C), Actinoseris (Fig. 5D), and Cnicothamnus
(Fig. 5E).
(2) Attenuate apiculate (Fig. 5G): in Gochnatia
secl. Leucomeris, hypoleuca, G. palosanto, G.
magna, and G. Е.
(3) Abruptly apiculate (Fig. 5J—L): in Gochnatia
sect. Discoseris, sect. Moquiniastrum (short append-
age in G. argentina, G. densicephala, and G. flori-
bunda), and sect. Pentaphorus, Gochnatia arbores-
сепз, ©. cordata, С. purpusit (Fig. 5J), Nouelia oe
5K), and Wunderlichia (very short) (Fig. 5L
Although useful in distinguishing taxa,
overlap exists between these broad categories such
м
some
as between apically caudate-attenuate (e.g.. lantho-
pappus, Fig. 5F), attenuate-abruptly apiculate (e.g..
Cyclolepis, Fig. 5H; Hyalis, Fig. 51), and abruptly
apiculate-caudate (e.g., G. velutina, Fig. 5J)
Anther tails are free and can be smooth or la-
They are smooth in the Caribbean species
—
ciniate.
of Gochnatia sect. Gochnatia (except G. attenuata,
G. ilicifolia, and G. microcephala), some species of
Gochnatia sect. Moquiniastrum (e.g., G. barrosit, С.
densicephala, G. floribunda, С. paniculata), some
species of Gochnatia sect. Hedraiophyllum (e.g.. G.
arborescens, G. magna, G. purpusii), Chucoa, Cni-
cothamnus, and Wunderlichia. The tails are lacini-
ate, at least in one side, in most species of Gochn-
alia (e.g., G. sect. Discoseris, G. sect. Leucomeris,
the South American species of G. sect. Gochnatia,
and G. glutinosa), and in Actinoseris, Cyclolepis,
Hyalis, lanthopappus, Nouelia, and Pleiotaxis.
STYLE
Most taxa have smooth styles except for Wunder-
lichia (Fig. 6A) and Chucoa (Fig. 6B) with dorsally
papillose styles, i.e., with the two branches (round-
ed and acute at the apex, respectively) covered by
short sweeping hairs, and Pleiotaxis (Fig. 6C) with
a subapical crown of short hairs. All species of
Gochnatia (Fig. 6D, E), Actinoseris (Fig. OF). Cni-
cothamnus (Fig. 6G), Cyclolepis (Fig. 6H), Hyalis
(Fig. 61), /anthopappus (Fig. 6J), and Nouelia (Fig.
OK) have smooth, apically rounded styles. Most
have the inner surface of the branches covered by
stigmatic papillae prolonged into the outer surface
of the style constituting a ridge of cells, which is
less evident in Cnicothamnus.
TRICHOMES
Leaf pubescence. Excluding some species such
as Gochnatia rotundifolia, Actinoseris hatschbachii,
А. stenophylla, and Hyalis lancifolia, which have
glabrous leaves, at least when mature, there are five
different types of trichomes in Gochnatia and its
relatives.
(1) Oblique septate flagellate hairs: one or two
foot cells, one- or more-celled stalks or stipes, and
unicellular, very long, flagellate, tubular heads (Fig.
7A). This trichome type is present in most species
of Gochnatia, Chucoa, Cnicothamnus, Pleiotaxis,
and Wunderlichia.
(2) Two-armed hairs (or T-shaped, malpighia-
ceous, anvil, dolabriform hairs): one or two foot
uniseriate,
and unicellular heads.
cells,
stalks,
which constitutes the head,
shape of a hammer and later becomes T-shaped by
further outgrowth of the two ends (Fig. 7B, С). This
trichome type is present in Gochnatia sect. Moqui-
niastrum (Fig. 7B) and G. cordata, Cyclolepis (Fig.
generally two- to more-celled
The apical cell,
initially assumes the
tE
— оо 690, LP). — Cnicothamnus lorentzii (Ruiz Leal —
Attenuate-apiculate anther appendage.
.L
Maung Mya 2
(Palacios & Cuezzo 2304, LP). —(
Cabrera (from left i to pied Gochnatia n ora
т — 3215, LP), G. palosanto (Schreiter in
(Joly . LP). —H.
ptly apiculate anthe : арре
amplexifolia (Hatschbach 35312, LP),
—
—
endage. J,,
Cyclolepis genistoides — 3172 Fa m LP. —
alia sect.
G. discoidea. — 3345, LP). J Go
Cabrera (from ie to right): ra — a (Hatschbach 9578,
(Rojas, herb. Hassler 9752,
maniana
paniculata (Gardner 4810, US), ie polymorpha (Glaziou in 1876,
64, ee ;ochnatia sect. Pentaphorus sensu Cabrera (from left to right): С
vernonioides (López
Р). G. purpusit (Pur — NY).
(Harley et al. 25209, MO) (very shortly apiculate).
‚б. oligocephala (Blanchet 328
. LP). J, (from left to right): С.
‚ Nouelia insignis (Maire 2
‘wo — Е. ake ен corymbosus
;ochnatia sect. comeris sensu
P), G. — ies 32, LP). G. G. hypoleuca
LP). G. — (Fernández 3666, NY), С. ат im
Hyalis lancifolia (Cabrera 4083, —
Discoseris sensu Cabrera (from ve jm EL
chnatia sect.
Р), G. les Ge — 5771 et al.,
8, US), б. orbiculata (Brade 5523, US), (
LP). G. velutina (Smith & Klein 14885, , 6.
>. fe оса
arborescens (Johnston 4023, LP), G cordata
516, NY). —L. Wunderlichia azulensis
538
Annals of the
Missouri Botanical Garden
A,B,D-F,H,1,K
0,3 mm
0,3 mm.
Figure 6. Style branches. A-(
—B. Chucoa ilicifolia (López Miranda 1090,
s. —D. Gochnatia discoidea nchet Be LP
(Hatschbach 28756, re —G. Cr
Kiesling 114, LP). —l. Hyalis argentea (Ruiz Leal 3701,
LP). —K. Nouelia — (Maire 2516, NY).
). —К.
7C), and Hyalis argentea. In Nouelia one end of the
apical cell is very short.
(3) Three- to 5-armed hairs (or 3- to 5-branched,
stellate hairs sensu Cabrera, 1971): similar to the
2-armed hairs, but the apical cell has З to
branches (Fig. 7D). The apical cell does not divide
and thus the head remains one-celled. This tvpe is
л
found only in Gochnatia barrosii and G. rusbyana
of section Moquiniastrum.
(4) Multistoried T-shaped hairs: similar to the
2-armed hairs, but further periclinal divisions take
place in the apical cell. The head is thus comprised
of 3 or 4 one-celled layers, all oriented transver-
sally and parallel, but of different lengths (Fig. 7E).
This type of trichome, not very common in Aster-
aceae, has been reported in the tribe Senecioneae
(Robinson, 1989). Janthopappus is the only genus
with this type of trichome.
(5) Biseriate glandular hairs: comprised of 2
rows of cells in the body, with two to many cells in
each row, enclosed by a persistent or collapsed cu-
ticular vesicle (Fig. TF, €).
Glandular hairs are
G. cordata аг — zi
¿ni — lorentzii n: Leal 14162,
e 2 branches. —4A. Wunderlic — azulensis (Harley et al. 25209, MO
Р). —C. Pleiotaxis eximia (Rolyns 1836, US).
D-K. Smooth style brane he
ctinoseris angustifolia
Р). — yck — Ы (Zardini &
Р). —J. к р Жозе. (Palacios & Cuezzo 2304,
widespread in all the taxa studied, and especially
in species of Gochnatia sect. Pentaphorus (i.e.,
Gochnatia foliolosa, С. glutinosa) where they cover
almost the entire surface of the leaf, with the fla-
gellate hairs restricted to the margins.
Achenial pubescence. Glabrous or slightly pa-
pillose achenes occur in Chucoa, in a few species
of Pleiotaxis (e.g., P. decipiens, P. huillensis, P. li-
nearifolia, P. welwitschii), and Wunderlichia (e.g..
W. crulsiana, W. insignis, W. senaei).
The remaining taxa have villose achenes.
M
W. bahiensis,
Three
types of achenial trichomes were found: Duplex or
twin hairs, leaf-like hairs (2-armed, oblique-septate
flagellate), and glandular hairs.
(1) Duplex or twin hairs: Twin hairs are the
common type in Asteraceae, comprised of two tri-
angular or rectangular basal cells, one sometimes
reduced, and two cylindrical or elliptical hair
cells, equal or subequal in length, generally in
contact up to their tips (Hess, 1938; Freire & Ka-
tinas, 1995) (Fig. 7H, 1). All the taxa studied, in-
Volume 89, Number 4
2002
Freire et al. 539
Gochnatia Morphology
E
$
B lo
=
E
E
wo
©
alo
C
=. €
B,C,E
0,1 mm
E
Е
5
| :
A
D FG
7 005mm | \\
0,1 mm. Y J K
Trichomes. A-G. Leaf hairs. —A. Oblique septate flagellate hair: d tortuensis (Ekman Н-3553.
Figure 7
S) E
3, C. 2- ~~ hair. —B. бос — — (Pedersen 8587, LP). —C.
Goc chnatia barrosit үсе th 16945, e
p n i) alacios & Cuezzo 2304, LP). К,
LP). —G. С — жем (Fabris e ы. H-L.
Nicora, LP). - to 5-armed hai
(Blanchet 3345.
lichia azulensts — el tal 25209, MO). H-I. I
hair. —K. Dup
Cowell 10183, NY).
cluding most species of Gochnatia, have achenes
with twin hairs, usually very long and filiform. In
some cases twin hairs have one hair cell very short
(e.g.. Gochnatia purpusii, G. recurva, G. tortuensis,
Clk Wunderlichia azulensis, W. mirabilis)
(Fig. 71), they are septate (e.g., Gochnatia hatsch-
bachii, G. oligocephala, Hyalis, Pletotaxis eximia,
Wunderlichia azulensis, W. mirabilis) (Fig. 7J), or
have only one hair cell (Wunderlichia azulensis)
(Fig. 7K).
(2) Leaf-like hairs:
achenes similar in morphology to the leaf hairs, as
Two-armed hairs in the
described above, were found in Gochnatia orbicu-
lata. Only a few species of Gochnatia (e.g.. G. cub-
ensis, G. magna, G. ramboi) have oblique-septate
flagellate hairs
(3) Ac heal biseriate glandular hairs: These are
Juplex hairs
yclolepis genistoides (Correa 3172 &
Multistor ed hair:
G. Biseriate каги уе н — К. Gochnatia o
Achenial hairs. H-k. Duplex c Wunder
the hair cells shorter tate dudes
ried T-sha
ith one of
ex hair with only one hair cell. —L. Capitate каб biseriate hair: бос — — (Britton &
similar to those on the leaves (Fig. 7F, G) and are
very widespread in the group under study, occur-
ring with the other types. A modification of the typ-
ical biseriate glandular hair with a very enlarged
head, i.e., capitate glandular hair (Metcalfe &
Chalk, 1950) (Fig. 7L). was found in the Caribbean
species of Gochnatia sect. Gochnatia.
PAPPUS
In all taxa analyzed the pappus is comprised of
rigid and scabrid bristles. However, there is inter-
esting variation in bristle length and width, modi-
fications of the lateral cells at the apex of the bris-
tles, and in the number of series of the bristles.
Five pappus types were found.
Type A. All the bristles have the same length
540 Annals of the
Missouri Botanical Garden
0,1 mm
Figure 8. ae General aspect (on the left) and detail of bristle apex (on the right). —A. Type A: Gochnatia recurva
E 20946, LP). —B. Type B: Gochnatia decora (Maung Mya 5309, LP). —C. Type C: Ge rchnatia cordata (Serrano 6, LP).
е D: Goc — и (Fabris 1343, LP). —E. Type E: Wunderlichia mirabilis (Irwin et al. 9913, NY)
and width (Fig. 8A). This is present in the Ca- Туре B. All the bristles have the same width,
ribbean species of Gochnatia sect. Gochnatia but about half are shorter than the others (Fig. 8B).
(except G. cubensis, G. oligantha, G. sagraeana, This type is present in Gochnatia sect. Leucomeris,
G. tortuensis), G. sect. Discoseris, and Actinoseris. the South American species of section Gochnatia,
Volume 89, Number 4
2002
Freire et al.
Gochnatia Morphology
most species of section Hedraiophyllum, and Chu-
coa, Hyalis, and Nouelia.
Туре C. АП the bristles have the same width,
half of them are shorter, and the longest are plu-
mose at the apex (Fig. 8C). This pappus is present
in Gochnatia cordata, G. sect. Moquiniastrum, and
lanthopappus.
Type D.
tively wide and flat (somewhat paleaceous) and the
Half of the bristles are long and rela-
other half are short and thin (Fig. 8D). This type is
present in Gochnatia sect. Pentaphorus and some
species of Pleiotaxis (e.g., P. dewevrei, P. eximia, Р.
pule — all with the long bristles acute at the
apex, and in Cnicothamnus, which has the long
bristles — at the apex.
Type E. Half of the bristles are long and rela-
tively wide and flat (somewhat paleaceous) and half
but the —
This kind «
are short and thin; all are scabrid,
are plumose at the apex (Fig. 8E).
pappus is found in Cyclolepis,
Pleiotaxis (e.g.. P. huillersii, P. rogersii), and Wun-
derlichia.
"
The type A pappus. i.e.
the same length, is 1-seriate with ЗО to 40 bristles.
Types B-E, i
are 2- to 3-seriate with more than 50 bristles.
e.. with the bristles of different length.
DISCUSSION
CHARACTER VALUE
Some characters such as habit, most leaf fea-
tures, and pappus vary within Gochnatia. Other
characters, although constant and common to all
species of Gochnatia, are not exclusive to it, such
as multiseriate involucres, tailed anthers with apic-
ulate anther appendages, and smooth style branch-
es. Table 2 shows this and the main morphological
characters that allow the genera associated with
Gochnatia to be distinguished. Some of the char-
acters studied merit a brief discussion.
The involucre of Gochnatia, resembling a spruce
cone, suggested by Hansen (1991) as a distinguish-
ing character for the group of Gochnatia and
associated genera, also occurs in other Mutisieae
such as Aphyllocladus, Dinoseris, Hyaloseris. and
Stenopadus, which are not very closely related to
Gochnatia. We do not therefore consider that this
feature has diagnostic value.
Corolla morphology has been used as a key di-
agnostic character in Mutisieae. The tubular acti-
nomorphic (Fig. 4A—E), tubular-filiform (Fig. 4G,).
tubular subzygomorphic (Fig. 4F,), subbilabiate
(Fig. 4H,. Т). and bilabiate corollas (Fig. 4J,. K,.
some species d
with all the bristles of
L,) present in the taxa studied show the great var-
Cabrera (1961,
1977) characterized Gochnatiinae and Barnadesi-
iation of this character in Mutisieae.
inae (the latter subtribe currently constitutes the
subfamily Barnadesioideae; Bremer & Jansen,
1992) in the first comprehensive key to subtribes
of Mutisieae by having more or less actinomorphic.
deeply 5-lobed disc corollas, with equal or unequal
segments, but never truly bilabiate, and bilabiate
or subligulate ray corollas. The Mutisiinae and Nas-
sauviinae, on the other hand, have clearly bilabiate
(the disc florets exceptionally actinomorphic) or lig-
ulate disc and ray florets. According to this key all
the taxa studied, although variable in their corollas,
would correspond to the subtribe Gochnatiinae
(sensu Cabrera, 1977). Other authors (e.g.. Robin-
son, 1991: Bremer, 1994) regard the distinction be-
tween Gochnatiinae and Mutisiinae, based only on
the actinomorphic versus bilabiate disc florets, to
be artificial and recognize only Mutisiinae sensu
lato.
Apiculate anther appendages have been strongly
considered to be an — character within Mu-
tisieae (Cabrera, 1977; Hansen, 1991; Karis et al.,
992: Bremer, 1994), The shape of the anther ap-
and abruptly
caudate, attenuate,
5C-L),
Gochnatia and also among genera. Indeed, Gochn-
pendage, i.e.,
apiculate (Fig. led to distinction. within
atia can be associated with Actinoseris. Cnicoth-
amnus, Cyclolepis, Hyalis, lanthopappus, Nouelia,
and Wunderlichia by the common possession of flat
(not bulbous), — anther appendages. Chucoa
and Pleiotaxis (Fig. 5А, B), based on their acute
and bulbous anther i f respectively, are
very different. The current tribal position of Pleio-
taxis is controversial. Some authors regard Pleio-
taxis as forming part of a mutisiean. “Dicoma-
group” (Bremer, 1994; Ortiz, 2000; Ortiz &
Coutinho, 2001). According to Hansen (1991), this
group of genera (including Pleiotaxis) should be ex-
cluded from Mutisieae by features of style branch-
es, anthers, and pollen type.
Smooth styles are revealed as another informa-
tive feature in this group of taxa, although the value
Some authors
of this character is controversial.
(Bremer, 1987) consider the glabrous styles to be
plesiomorphic within Asteraceae (although the Lob-
eliaceae, characterized by hairy style branches,
were used as an outgroup), while others (Stuessy et
al.. 1996) postulate the smooth condition to be de-
rived with Calyceraceae as the outgroup. At present
Calyceraceae are the sister
2001; Ur-
tubey & Stuessy, 2001) is widely accepted. Since
the hypothesis that
group of Asteraceae (e.g., Albach et al.,
the style branches in Calyceraceae are papillose
Annals of the
Missouri Botanical Garden
qo
542
J 9d Я ‘q sedA] g ә44], 7) ad Ay g eday, 4 әчАү, q edA] g oda], y әд], ‘q “y sadA] snddeg
ÁI
A[[esıop -[eoıdeqns Аеѕлор
пиәәѕәфпа пиэәѕәфпа yyoours yjoous q10ouis u1oouis yloours :u22saqnd yyoous yjoouls — soqouedq IJÁIS
IPUI
oous 3]eruroe| әүрїшәР] 3]eruroe| 3peiuroe| apeiuroe| оош» yjoous әүрїшәР] -вү "[joours sre dou y
asoq ayenuaye
-mq *pauo ayenuaye ЕТЕ ayenuaye ‘apne?
храпе -Jr Apdnaqe -9]ppneo -Apidniqe -Andruqe ayepneo ayepneo ‘храпе osepuodde
:9je[noirde :9je[noirde :9je[norde :9]e[norde :3]e¡norde :9]e ¡norde :aye¡norde əmər :9je[norde ap pnaido porde тәщиұ
(әицм
WERI *ogue
“AM a[dınd a[dınd -10 °2P[1[)
MOT[SÁ “OPIUM ‘yurd “pal IJYM pue әйцм әр MOT[OA IBUBIO моәл “уша #әйум — швәлә “морәл 10[02 P[[O107)
(orqdaour
-Ipqns)
orgdaouiost orqd.ouiost ornqdaouip orqd.rouirp omjdiounp — orngdaouitpqns ongdaourp orydaowost orqd.rouip nydiowosi se[[0407)
3184229] agepuadde
juasqe juasqe juasqe juasqe juasqe juasqe “SIQUIOYA juasqe juasqe juasqe ¡eorde sourep sug
(snoui snoul (snow
-e30.19194) -e3019194 -e3019194)
snowersowoy snoureaoulou snowe3owoy —snourm2019]3Qq snourezourou snouresouloq snowegowoy snoure3ouoy ‘snowesowoy snowpsowoy epnuder,
91eo[ed әүвә[едә әвә[вЧә әвәрә 3jeo[eda ajeo[edo a1e2[eda 3jeo[edo 3jeo[pdo әүвә[еаә әүәруЧәдән
pourre
-¢ 01 зәшоцә
padeys-J -g 'peuue = 1210)
aje[ozep aje[[o2eg pouue-z — porrojsnnur рәшие-2 pauLie-z 31R[[93P y a1P[[938y aje|[22eg -Z wm[pSeg ^ 1e[pnpue[32-uoN
sqniys
snoyAyde (sqnuusqns)
saad] sqiou sqnaus sqniysqns sqniysqns *asouids s331} sqnays sqıay *sqnuys ‘s22. I qeu
Diyana puny sixDjo19]d р1]әпо\/ snddodoyrun] sig s1d210]2&7) snuuruJ]1021u7) poony’) suasouioy D11muwu2o-)
“SNJN ur әле P19Uu93 13410 шол) DIJDUYIOL) ysin3unsip ey} S19]JB.IB Uy”) "sasagjuaaed
€ 1
-
=
I әле $31n]£9J UOUIUIODU N '99€jp[oq UL ӘЛЕ SI9]OETIEQO 3AIsn[oxa эц], "P19u238 pare 1dosse pue DIJDUYIOL) jo uoroutrjstp a41 мо[е ey} SIIJOPIBYO [e2130[oydıou шеу
Volume 89, Number 4 Freire et al. 543
2002 Gochnatia Morphology
(Hansen, 1992, and our own observations) the sect. Gochnatia, sect. Hedraiophyllum, and sect.
smooth style branches in Asteraceae become an ad-
vanced character for the family.
The multistoried T-shaped hair (Fig. 7E). the 2-.
and 3- to 5-armed hairs (Fig. 7B-D), the capitate
glandular achenial hair (Fig. 7L), and the pappus
types established here (Fig. 8) are revealed as new
diagnostic characters. In fact: (1) the multistoried
T-shaped hair is exclusive to lanthopappus and be-
the
2-armed hairs are present in Gochnatia cordata, G.
sect. —— Cyclolepis, Hyalis, and Noue-
lia; (3) t to 5-armed hairs are exclusive to
some species of Gochnatia sect. Moquiniastrum; (4)
the capitate glandular achenial hairs, with a very
enlarged head, distinguish most of the Caribbean
comes another character to distinguish it: (2)
the :
species of Gochnatia sect. Gochnatia; and (5) pap-
pus types A, B, C, D are present in Gochnatia and
allow the distinction of sections within the genus
(see below), while type E is present only in Cyclo-
lepis, Pleiotaxis, and Wunderlichia.
CIRCUMSCRIPTION OF GOCHNATIA AND ITS SECTIONS
Our morphological analysis confirms that Gochn-
atia has no single exclusive feature that distin-
guishes it from related taxa, but it can be defined
by a set of characters, i.e., homogamous (rarely het-
erogamous) capitula, isomorphic (tubular) to sub-
dimorphic (tubular and tubular subzygomorphic)
corollas, apiculate anther appendages, and smooth
style branches (Table 2).
After analyzing the morphological features of
the sections established by Cabrera
(1971) were reviewed and some changes are pro-
posed
Gochnatia,
The distinctiveness of two of the six Cabrera sec-
tions, sect. Discoseris and sect. Pentaphorus, was
confirmed and additional characters supporting
them were found. For instance, Discoseris has pap-
pus type A and Pentaphorus type D.
Gochnatia cordata, placed by Cabrera in section
Hedraiophyllum, has characters that link it to sec-
tion Moquiniastrum such as the 2-armed foliar hairs
(Fig. 7B), numerous capitula arranged in glomeru-
lose pseudopanicles (Fig. 2H), and pappus type C
(Fig. i
However, it constitutes the type species of section
8C) and must be included in this section.
Hedraiophyllum established by Lessing (as a sub-
genus) in 1832 and has priority over the name of
section Moquiniastrum established by Cabrera in
1971 (with G. polymorpha as the type species).
Thus, the name Hedraiophyllum is retained in what
was known until now as section Moquiniastrum.
Phe remaining three sections sensu Cabrera,
Leucomeris, are redefined resulting in the establish-
ment of five sections: sect.
Gochnatia,
sect. Rotundifolia.
Cabrera (1971) established two groups in his key
to section Gochnatia, the “South American species”
and the
Anastraphioides, sect.
Glomerata, sect. sect. Leucomeris, and
“Caribbean species.” Species of this sec-
tion have some characters in common, such as sol-
itary or 2 to 3 capitula and caudate anther ap-
pendages, suggest that it
should be separated into two sections: sect. Anas-
traphioides and sect.
but other characters
Gochnatia. The Caribbean
species form the new section Anastraphioides clear-
ly differentiated by spiny leaf margins (Fig. 1E).
corollas very deeply 5-lobed (Fig. 4B), and pappus
type A (Fig. 8A). The South American species, on
the other hand, have predominantly ovate leaves
with entire margins (Fig. 1A, F), corollas deeply
2-lobed, and pappus type B (Fig. 8B), and corre-
spond to section Gochnatia.
The species Gochnatia rotundifolia, included by
Cabrera in section Gochnatia, has characters that
show a departure from the other sections of the ge-
nus such as glabrous and 3-veined leaves, white
corollas, and anther appendages attenuate. It would
approach Cabrera’s section Leucomeris by its phyl-
laries glabrous with ciliate margins and conspicu-
ous parallel veins, but G. rotundifolia lacks the in-
volucre extending into the peduncle typical of this
section. Consequently, this species is placed in the
new, monotypic section Rotundifolia.
Gochnatia sect. Hedraiophyllum sensu Cabrera,
which was already recognized by Cabrera (1971) to
be artificial, is split off. Three species of this sec-
tion, G. arborescens, G. magna, and б. purpusti, are
placed in the new section Glomerata characterized
by the exclusive presence of numerous capitula ar-
ranged in glomerulose pseudocorymbs (Fig. 2C). As
mentioned above, the type species of section Hed-
raiophyllum, G. cordata, was moved to Cabrera’s
Gochnatia sect. Moquiniastrum. The remaining spe-
cies of section Hedraiophyllum, G. hypoleuca, G.
palosanto, and G. smithii, were placed in section
Leucomeris largely based on their involucre with
glabrous, conspicuously veined phyllaries, ciliolate
al the margins, extending into the peduncle (Fig.
31).
In summary, we propose the following eight sec-
tions for the genus Gochnatia: sect. Anastraphioi-
des,
atta, sect. Hedraiophyllum, sect. Leucomeris,
sect. Discoseris, sect. Glomerata, sect. Gochn-
sect.
Pentaphorus, and sect. Rotundifolia. This new in-
frageneric classification, with descriptions and а
key to the sections, is shown in Appendix 2.
Annals of the
Missouri Botanical Garden
544
DIJDUYIO*) UTYJIM suoroos osje pue “BXB] pa1e]91 ғи шолу DIJDUYIO) JO uonoutjstp əy} зшмо[е *xo[duioo DIJDUYIO*) 34] suluyap SIIJIBIB YO jeorsojoydiow шеу
'uJuojideos ejnydeo t
'6 210314
'ejejnounped uoj squiÁsodopnasd =.
asojnjawo¡B squiÁjooopnesd | ‘ец рәшше-ому x 'esouids uiDueui je a]
pi | vile
`2 adA} sndded Т ‘а edA sndded mm “y addy sndded ПП 'ejeuo' деб sauej¡Ayd f] 'eyeues 2}-/ Ajjuquedde “mh
| 1
ѕпошебошоц A
ѕпошебвојәјәц
snddedoujue|
sepioiudejseuy
SUM JO 38]!|
yes
sejjoJo9 \
snsoydejuag |
19S BJEIBUO/O “pas (|!
$ШӘ$025!0] Pas
ецеицооо pas
BIOJPUNJOY pas SYM
SE||O0J02
ıeıpaH
es
suewoone? Pas
xejduio2 ецеиц2оо
Volume 89, Number 4
2002
Freire eta
: 545
Gochnatia Morphology
THE GOCHNATIA COMPLEX
The combination of apiculate anther appendages
and smooth style branches is exclusive to Actinos-
eris, Cnicothamnus, Cyclolepis, Gochnatia, Hyalis,
lanthopappus, and Nouelia within Mutisieae. Con-
sequently, these genera are recognized here as the
Gochnatia complex (Fig. 9). Chucoa, Pleiotaxis,
and Wunderlichia, as already mentioned, differ in
their styles and anthers.
Within the Gochnatia complex, the genus most
morphologically similar to Gochnatia is Cyclolepis.
This genus shares trichomes 2-armed, subdimorph-
ic corollas, and gynodioecy with section Hedraio-
phyllum, which constitutes a pivotal group among
the remaining species of Gochnatia and Cyclolepis,
providing a “link” between both genera. Cyclolepis,
however, can be distinguished from Gochnatia by
its leafless, spiny branches, tubular, filiform female
florets (Fig. 4G,), and pappus type E (Fig. 8E).
The hypothesis that a shrubby habit, inde ho-
mogamous, and solitary or few — with yellow,
and actinomorphic corollas represent primitive con-
ditions in Asteraceae has been widely discussed
and recognized (Maguire & Wurdack, 1957; Ca-
rlquist, 1976; Bremer, 1987, 1994; Pruski, 1991:
Stuessy et al., 1996). From this point of view, most
species of Gochnatia have a set of plesiomorphic
characters when compared with the remaining gen-
era of the complex. This hypothesis is consistent
with Cabrera's (1977) idea that Gochnatia is the
basal genus in a complex from which the other gen-
era in the subtribe are derived.
Literature Cited
Albach, D. C., P. S. Soltis, D. E. Soltis & R. G. Olmstead.
2001. Phylogenetic analysis of Asterids based on se-
к of four genes. Ann. Missouri Bot. Gard. 88:
3-212.
© G. M. & В. Maguire. 1973. A review of the genus
Wunderlichia (Mutisieae, Compositae). Rev. Brasil. Biol.
33(3): 379-406.
Beltrán, п. — R. Ferreyra. 2001. Una especie nueva de
e para ey y Boli id Gochnatia lanceolata.
C — Newsl. :
Bremer, K. 1987. — Pn of the Astera-
ceae. Cladistics 3: 210-253.
994. Asteraceae. Cladistics and Classification.
Timber Press, Portland, Oregon.
& R. K. Jansen. 1992. A new шуу" of As-
leraceae. Ann. Missouri Bot. Gard.
Cabrera, А. L. 1950. Observaciones ma p géneros
Gochnatia y Moquinia. Notas Mus. La Plata, Bot.
15(74): 37-48.
. 1951. Notas sobre — de América Aus-
tral. Darwiniana 9(3—4): : 386.
1955. Un nuevo э 66 nero de Mutisieas del Perú.
— Bot. 6: 40-44.
PL argentinas: Clave para la de-
Bol. Soc.
e de los géneros. Revista
, Bernardino Rivadavia Inst. Nac.
Bot. 2: 291-362.
. 1970, Actinoseris, nuevo género de compuestas.
Bol. Boc. Argent. Bot. 13: 45—52.
1971. Revisión del género Gochnatia E neon
tae). Revista Mus. La Plata. Bot. 12(66): 1-16
974. Tres Compositae nuevas de Minas Gerais
(Brasil) Mus. Bot. Munic. 15: 1—4.
— . Mutisieae. Syateriatis review. Pp. 1039—
1066 i in v b Heywood, J. B. Harborne & B. L. Turner
(editors), The Biology a SUME of the Compositae
Academic Press, Loi
Candolle, A. P. de. 1838. —— 7(1): 24. Par
Carlquist, S. 1976. Tribal Ln and м УВ
ny of the Asteraceae. Aliso 8: 465-49
Font — P. 1979. Diccionario de —* Ed. Labor,
S. A., Barcelona, España.
Franc he M. A. 1888.
Bot. 5: 65-67.
Freire, s E. & L. Katinas. 1995. Morphology and ontog-
eny of the cypsela hairs of Nassauviinae (Asteraceae,
Mutisieae). Pp. 107—143 in D. J. N. Hind, C. Jeffrey &
G. Pope (editors), Advances in | Compositae Systematics.
(ew.
Mus. Argent. Ci.
Invest. Ci. Nat.
Les Mutisiacées du Yun-nan. J.
Phe ореви oh in Compositae
tribe Vui a Bot. 109:
)92. Studies in the € aceae with a discus-
sion a its relationship to C ae Nordic J. Bot. 12:
63-75.
Harris, J. ( M. W. Harris. 1994. Plant Identification
Terminology: An i. Glossary. Spring wake Pub-
lishing, Spring Lake, Utah.
Hess, R. 1938. Vergleic — Untersuchungen über die
ТРЕ ИЕА der Kompositen. Bot. Jahrb. Syst. 68:
135—496.
Jansen, K. K.-J. Kim. 1996. Implications of chlo-
roplast DNA E for the classific: ve and е
of the Asteraceae. Pp. 317-339 in D. J. N. Hind & H
J. Beentje (editors), Compositae: — лас. Proceed-
ings of the DE e Compositae Conference, Kew
1994, A . The Royal Botanic Gardens, Kew.
d Palmer. 1987. A chloroplast DNA inver-
sion bs an ancient pound split in En я
. 84:
ег — (Asteraceae). Proc. Natl. Acad.
5818-5822
Jeffrey, c — Notes on с, a The сне
in East M Africa. Kew Bull. 1
Jervis, R. N. 1954. A Summary of E РЯ
including а Revision of the West Indian — which
Comprise the Sec tion е du Ph.D. Disserta-
tion, e of Michigan, Ann Arbo
Karis, O.. M. Kallersjö & K. Bremer. 1992. Phylogenetic
— sis Е the Cichorioideae (Asteraceae), with empha-
s on the Mutisieae. Ann. Missouri Bot. Gard. 79: 421—
===, Р, e & M. Killersjó. 2001. New evidence
for the systematic position of Gundelia L. with notes on
delimitation of Arctoteae (Asteraceae). Taxon 50: 105
114
Kunth, C. 5. 1818. In F. W. ips A. J. A. Bonpland
& C. S. Kunth, Nov. gen. sp. 4
Реа С. Е. 1830. Synanthereis таа Regii Berli-
nensis, Dissertatio Tertia. Linnaea 5: 237-263.
. I6 e лана Syn. Gen. Compos.: 102-10
«de
Mais, B. Wurdack. 1957. The botany of
546
Annals of the
Missouri Botanical Garden
Guayana — Part Il. Mem. New York Bot. Gard.
93): 235-3
Metcalfe, C. R. & L. Chalk. 1950. Anatomy of the Dicot-
— Leaves, Stem, and Wood in Relation to Tax-
onomy, with Notes on Economic Uses, Vol. П. Claren-
don Press, Oxford.
Moreno, N. P. 1984. Glosario Botánico "yen
pañía xw Continental, S. A. de C. V. (C.E.C.S
Xalapa, Mexic
Ortiz, S. 2000. A phylogenetic analysis of Dicoma d
and related genera (A
ieae) based on morphological and anatomic characters.
Ann. Missouri Bot. Gard. 87: 459-481.
& A. P. Coutinho. 2001. Achyrothalamus reduced
to гуен (Asteraceae, Mutisieae). Taxon 50:
389—40:
Com-
. A
—
Р
steraceae: Cichorioideae: Mut
Payne, W. W. 1978. A pon of plant hair terminology.
Brittonia 30: 239-255
Pruski, J. К. 1991. — of the Guayana High-
* I—V. The Mutisieae of the Lost World of Brazil,
C olombia and Guyana. Bol. Mus. Paraense, ser. Bot. 7
335-392.
1962. Studies in the trichomes of some Com-
India 4:
Ramayya, N.
e — M General structure. Bull. Bot. Surv.
-18
— A 1989. A revision of the genus Dresslero-
thamnus (Asteraceae: Senecioneae). Syst. Bot. 14: 380-
1991. Two new species of Stifftia with notes on
lationships of genus (Asteraceae: Mutisieae). Sys
‚ 16: 685-69
Roque, N. 1
(Less.) Cabrera (C —
eS Kew Bull. 52: 1
. Five new species of Richterago (Compos-
itae, ы A genus endemic to Brazil. Novon 11:
341-34 a
re
997. ie reassessment of Actinoseris polymorpha
Mutisieae) и а пем
97-204.
=
& . М. Hind. 2001. /anthopappus, a new ge-
nus of us tibe Mutisieae (Compositae). Novon 11: 97—
101.
—— — & J. R. Pirani. 1997. Flora da Serra do —
Minas Gerais: Co WO NT SET ae e Mutis
35.
Bol. Bot., Univ. Sáo dapes x 151—1&
& . 2001. Re — ment of the name Ri-
chterago — жер recircumscription of the genus to
include species formerly treated as Actinoseris (Endl.)
Cabrera (Compositae, Mutisieae). 1155-
160.
Sancho, G. 1997. Revisión Sistemática, Análisis Cladís-
tico y Biogeográfic o de la Se
Taxon 50:
ción И del
a (Asteraceae, Mutisieae). Ph.D. Dis-
Universidad Nacional de la Plata, La Plata,
Género Gochnati
sertation,
Argentina.
————, 2000. Revisión y filogenia de la sección Moqui-
niastrum Cabrera del género Gochnatia Kunth (Aster-
‚ Mutisieae). Колин 54(5): 61-122.
Stuessy, T. F., T. Sang & M. L. 1990.
and locura of the Barnadesioideae, with impli-
cations for early о м Compositae. Pp. 463—190
in D. J. N. Hind & H. J. Bee ntje (editors). e
Systematics. Proceedings p: the e Compos-
itae Conference, Kew 1994, Vol. 1. The Royal Botanic
w
Devore. Phylogeny
—
T
€.
Urtubey, E. & T. F. Stuessy. 2001. New hypotheses of
\ I
phylogenetic relationships in Barnadesioideae (Astera-
3
ceae) based on morphology. Taxon 50: 1043-1063.
Zardini, К. М. 1975. A new species of | Actinoseris Com-
positae). Bol. мы Bot. Munic. 23: 1—
—
APPENDIX 1.
Index to specimens examined, with vouchers. Note:
During preparation of this work, the new species Gochn-
atia. jd eolata Beltrán & Ferreyra was published (Beltrán
erreyra
(C 'arolina 01. журе US;
2
"zj
)1). Spec — rn pe п! species
о & Peca 336, US) have
\ асще enis appendages, so this
species must be excluded from Gochnatia and does not
belong to the Gochnatia complex
Actinoseris
А. ек BRAZIL. Minas Gerais: Jaboticatuba,
schbach 28756 (LP); Serra do Cipó, Hatschbach
986 5 (LP).
A. утара: BRAZIL.
ade 7 (LP).
А. — "BRAZIL . Minas Gerais: Santa Ana do Ria-
cho, Hatsc on h 35304 (LP).
A. radiata: a а a Paraná: Campo Largo, Hatschbach
690
Minas Gerais: Serra de Cipó,
А. илину: Minas Gerais: Santa Ana do
iacho, de Tm т; 35388 & Koczicki (LP).
Cyclolepis
C. La PARAGUAY. Chaco dea Rojas 7104
(LP). GENTINA. — bas | de las Con-
chas, us 3782 (LP). — Gral. Cone-
ssa, Correa 3172 & — (LP). Salta: La Cande-
laria, Schreiter 6630 (LP). San Juan: La Laja, Tinto
(LP). Neuquén: Plotier, Banda, Zardini &
Kiesling 114 (LP).
Chucoa
C. ilic jm PERU. La T dia Santiago de Chuco, Ló-
z Miranda 1090 (LP).
Cnicothamnus
C. azafran: ARGENTINA. Salta: Serranías del Crestón,
Bortagaray 22 d Ab P). Jujuy: Santa Bárbara, Cabrera
et al. 26287
C. lorentzii: ARG AT г INA. Salta: Caldera, Cabrera et al.
Capital, Bonavía 61 (LP); camino de cornisa Salta-
Jujuy, Zardini 1292 (LP).
Gochnatia
G. amplexifolia: BRAZIL. Minas Gerais:
Riacho ao 35312 (LP).
G. arborescens: MEXICO. Baja California: Isla Cerralvo,
— 4023 (L P: entre Santonio y Puerto de Ba-
» los Muertos, Wiggins 5632 (US); along the
^ac ей Coast, 14 mi. S of Pescadero, Spjut & Edson
)85 (US).
Santa Ana do
`P
ж
(
С. gio PERU. Arequipa: Monte Chiwata, Eyer-
dam & Beetle 22120 (LP).
G. Кошто Tucumán: Trancas, Meyer
— (LP). Entre Rios: Paracao, Paraná, Schulz
3 (LP): Leales, vanduti 703 (LP).
G. argyrea: BRAZIL. Paraná: Vila Velha, Dusén
(LP), Dusén 9115 (G), Hatschbach 9578 (LP)
G. attenuata: CUBA. Oriente: Sierra de Nipe, Ekman
19174 (1
4035
Volume 89, Number 4
2002
Freire et al.
Gochnatia Morphology
~
м 2
~
a
c
a
EN
`
c
a
см
$
ea,
M
м
ч
~
`
"s
ч
~
i
c
a
м
`
ы 0 BOLIVIA. Santa C
М бшем:
. calcio:
М — nasil:
€ —
re crassifolia: CUBA. Oriente:
>. cubensis
. CUTU — BOLIVIA. Tarija:
. de nsicephala: BRAZIL.
d divider
r dise — ВК,
,. ше»
. in
. foliolosa:
345
М gardneri: BRAZIL.
r. —
LP
p. gouni К i BA.
.. hatsc nid hit:
n» haumaniana:
. — MEXIC(
Hatschbach
Macedo 1138
Paraná: Cianorte,
serais: [tuitaba,
or Tros E i:
BRAZIL.
5 (LP). Min
Minas
ia a PARAGUAY. Amambay: Parque Nacional
¿erro Cord, Sancho 8 (LP).
blanchetiana: BRAZIL. Ceará: Serra do лк
;ardner 1735 (K). Goiás: Glaziou 21663 (G
ruz: Florida, — &
Gutiérrez 33804 (LP).
DOMINICAN REPUBLIC.
(LP
nte Cristo, Jiménez 1598 (LP).
IB
Falda del Morro,
Oriente: Guantánamo, Ekman
64 (S).
ARGENTINA. Jujuy: Tafna, Cordo & Fer-
rer 88-B-17 (SD.
ARGENTINA. Entre Ríos: Federación, Bur-
3169 & Crespo (LP): Santa Ana, Serrano 6
(L P. T AZIL. Rio Grande do Sul: Porto Alegre,
Rambo 545 (LP).
.. cowellii: CUBA. Santa Clara: Santa Clara, Britton «e
Cowell 10183 (NY
Baracoa, Ekman 4023 (S).
: CUBA. Oriente: Sierra de Nipe. Ekman
4767 (S). 9632 (S).
— Fiebrig 2838 (LP).
¿NTINA. Jujuy: Humahuaca, Meyer 21. id
Д n p La Cande a Schreiter 9409 (LP): La
ndelaria, Jeréz et al. 49120 (LP
. 2 MYANMAR. — Lweji, Maung Mya 5309
(LP).
Rio de Janeiro: Rio de Ja-
neiro, Glaziou 11072 (K). Minas Gerais: Pocos de
— Leoncini 88 (LP): Pico da Bandeira, Shep-
5771 et al. (UEC).
BRAZIL.
5 (LP).
Bahia:
Igregia Velha, Blanchet
AZIL. Minas Gerais: Claussen 1301 (NY ).
J.
A. Sin. loc., Wright 2875 (GH).
—— AN REPUBLIC. Cordillera E en-
tral, Samana, Los Haitises, Ekman H-15498
Minas Gerais: Morro d Frfo,
: Diamantina, Hatschbach 30192
istrito Fede ral: Brasilia, Hatschbach 43151
ardner
(US). D
Y
CHILE. Santiago:
Cerro Renca, Cabrera
Valparaíso: i
51 (LP). El Quisco, Маһи 10336
Goyaz: Capella da Passe. Gardner
).
ARGENTINA. Tucumán: Tafí. Fabris 1343
ndoza: Cerro de La Gloria, King 183 (LP).
Oriente: Región de Moa, Cerro Mir-
. Marie-Victorin et al. 21591 (GH).
BRAZIL. Minas Gerais: Jabotic atubas,
Hatschbach 29951 (LP); Datas, pr hbach 301.
LP).
aflore
BRAZIL. Mato Grosso do
Porá, Meyer 18770 (LP). PARAGUAY.
Sierra de — Rojas 6575 (LP)
— 9752| (G)
1 (L
Sul:
Amambay:
). Rojas [herb.
^st. Los cinco hermanos, Sancho
Ponta
). Hidalgo: С
;onzález Quintero 3215 5 (LP).
Р).
Cañada del Vaquero,
Coahuila: Sierra Gav-
а. Johnston 7223
А ilicifolia: BAHAMAS ISLANDS. Andros Island: Cop-
pice, Small & Carter 8526 (К)
G.
G.
G.
G.
G.
~
`
~
Ss
re кш
. microcephala: CUBA. Oriente:
NY
: — BRA
А тота па:
. E. igocephala:
. iris ata:
pare ili CUI
38 (NY),
. pauc for ulosa:
А — BRAZIL.
a —
. recurva:
rusbyana: BOLIVIA. Yungas: В
. sagraeana: CUBA.
~ Shaferi: CUBA. Oriente:
ñ —
4 us tabilis:
intertexta: CUBA. Pinar del Río: Cajalbana, Alain A-
1680 (NY
magna: MEXICO. San Luis Potosí: Cronquist 11277
(NY): Queretaro, 5 km SW Jalpan, Fernández 3666
NY).
CUBA. Pinar del Río:
8 (LP).
Guane, Shafer
Boca Guantánamo a
itua, Bro. — 4874
a Rio Grande dis Sul: Malme 646
5); pr. 3 Santa aah Malme 1261 (5).
A. Pinar del Río: Guane. Ekman
5 (S).
e е СОВА. juger 2:4 тп Baracoa region,
NY).
sa de Prada, /ебп 119
e ligar REPUBLICA eh AN, i Monte Cristi,
of Villa Isabel, Jiménez 3614 (US
BRAZIL. Bahia: — ra ae Jacobina,
Blanchet 3288 (U: S); Río Branco, Curran 284 (NY),
Salzmann s.n.
= ВК, AZIL. Sao Paulo: Moóca. Brade 5523
S). Rio Grande do Sul: Guaiba. Sancho 48 (LP):
e ше y poe — 998 (К).
—
С”)
T
y
palosanto: ARG A. Jujuy: San Pedro, Cabrera
F — ee үз E Tucumán: Vipos, Trancas,
Venturi 1296 (LP), Schreiter in 1925 (LP).
BRAZIL. Minas е rais:
Р), Gardner 4810
JA. Oriente:
Massa. Brade
а 'n Savannas, Shafer
. pain PERU. La Libertad: Pataz entre Huaylillas
‘A
Pavabamba, López & Sagdstegut 3409 (LP).
BAHAMAS NDS. Fortune Is-
land: Eggers 3866 (К); Mariguana Island, 10 mi. W
of Abraham Bay, Wilson 7.
picardae: HAITI. , eroix-des Bau-
quets, gorge of Grande-Riviére de cul-de sac. Ekman
Н-5385 (К).
—
Massif de la Selle
М — BRAZIL. Rio de Janeiro: Tijuca, Glaziou
6 (LP). Sao Paulo: Paranaiba do Sul, Hash-
imoto 624 (LP). Rio Grande do Sul: Pereira 8609
& Pabst 7984 (LP). PARAGUAY. Amambay: San-
es 26, Sancho 32 (LP). San Pedro: Lima. Pedersen
(LP).
Sao Paulo: Campos de Emas, Ca-
brera 12311 (LP); Ityrapina, белг 8296 (LP).
—
. purpustt: MEXICO, Puebla: Tehuacán, Purpus 4248
(NY).
BRAZIL.
> (5).
— e p)
CUBA. Oriente:
946 (LP).
Xanxeré, Rambo
Palmeira. Rambo
Santa Catarina:
Rio Grande do Sul:
entre Moa y Punta Andén,
León 20
, rotundifolia: BRAZIL. Sao Paulo: Capital. a
Handro 157 (LP); Vila Esperanga, Joly 59
Alto Urumbamba, Zamalloa 2015 (LP).
Habana: León 7094 (NY); Vedado,
Cusco:
Main 2532 (NY).
La Caridad, López Figueiras
1738 3 (NY).
М шт MEXICO, Oaxaca: Cuesta de Coyula, Conzatti
5 (US).
BRAZIL. Parana: Ponta por Hatschbach
(LP); Senges, M hbach 2 K).
INDIA. tar Pradesh:
)un District, Calrola 32 (LP).
7422
ea Dehra
548
Annals of the
cae Botanical Garden
G. — HAITI.
"aix, Ekman H-355.
G. vargasii: PERU. — Abancay, Vargas 16317
(LP).
C. — BRAZIL.
Hatschbach 23447 (LP),
lein 14885 (LP).
G. iii hat Re RU. Amazonas: Chachapoyas, Tingo,
— 97 (LP 2 e i Chachapoyas y Leimebam-
ez еі al 1 (LP). La Libertad: Bolívar,
— 1701 (L E ps & Sagástegui 3354 (LP).
— du Nord-Ouest, Port-de
т
Parana: Рома Grossa, Vila Velha
López 4364 (LP), Smith &
Hyalis
H. —
еа (LP).
Н. lane ifolia: ARGENTIN haco: San Fernando, Ca-
brera 4083 (LP). Sc — 16098 (LP).
ARGENTINA. Mendoza: Tupungato, Ruiz
701 (LP
lanthopappus
f, us aca ARGENTINA. Corrientes: Paso Troncón,
'alacios & Cuezzo 2304 (LP).
Nouelia
N. insignis: CHINA. Maire 2516 (NY). Yunnan: l’ Abbe
Delavay 2498 (US), Rock 11714 (US).
Pleiotaxis
P. dewevrei: DEMOCRATIC
kulu, de Hitte 288 (U$
P. eximia: DEMOCRATIC REPUBLIC OF CONGO.
Tshinloingwe (Hout Katanga), Rolyns 1836 (US).
ZIMBABWE. 20 km de Mangula, Lavranos 22745
REPUBLIC OF CONGO. Lu-
S).
(US).
P. huillensis: ANGOLA. Gossweiler
10780 (US).
P. pulcherrima: ANGOLA.
Gossweiler — US
Р, je ide ZAIRE—DE MOC RATIC REPUBLIC O
CONGO. E Tem thville, Rolyns 1568 (US).
Huila, Hampata,
Distr. do Cuanza Sul, Seles,
`
K
Wunderlichia
W. yes nsis: BRAZIL.
l. 25209 (MO).
W. vilia BRAZIL. Goiás: Chapada dos Veadeiros,
ter et al. 2615 (MO).
W. —— BR) AZIL. Goiás: Serra do Cristais, Irwin et
al. 9913 (NY); about 52 km W of Alto Paraiso, Mar-
tinelli & Stutts 999 (NY
edra Azul, Harley
Minas Gerais: |
~
ч
APPENDIX 2.
New infrageneric classification of Gochnatia.
KEY TO THE SECTIONS OF GOCHNATIA
l. Pappus type A (all bristles are thin and have
the same length and width) and/or pappus type
B (all bristles are thin and have the same width,
but about half are shorter)
C (all bristles are thin and have
~ and
N
the longer are plumose at the apex) or pappus
type D (half of the bristles are long and wide,
and the other half are short and thin)
2(1). Subshrubs; capitula solitary, very long-pedun-
culate, or Wan in se apiform pse ges 'orymbs
G. . Discoseris
~
24. Tres or shrubs; capitula short-pedunculate or
essile D
3(2). кубе phyllaries dorsally oo or
subglabrous and ciliolate at the marg
3'. Involucral phyllaries dorsally tomentose and.
not ciliolate at the ma
3). Capitula arranged in glomerulose pse eudoco-
Seu еа les:
TAINS ә
ÉS
RE
rymbs, pseudoracemes, or |
phyllarie s extending into the pedun
G. ‚ Leucomeris
4. Solitary capitula; — not «ш ы into
the peduncle G. sect. ni de.
5(3). Solitary capit tula o r2or3
5. € apitula i in glomerulose — 'orymbs
G. sect. Glomerata
ly
6(5). Leaves spiny; corollas deeply to very deeply
lobe t. Anastraphioides
O”. Leaves not spiny; corollas des Arn Tobe d
G. sect. Gochnatia
Pappus type ( 2: :; plants not glandulose, with :
and 3- to 5-armed hairs; gynodioecious or po-
lygamous бес ious: corolli as subdimorphic
(isomorphic), 9-15(50) ; sect. ا
7'. Pappus type D; plants | gl Caes monoecious
corollas isomorophic, 5 (3, 7, or 10-20
7(1).
~
A
ER sect. Pentaphorus
Gochnatia sect. Anastraphioides Jervis ex S. E. Freire,
„ Katinas & G. Sancho, sect. nov. TYPE: Gochnatia
ilicifolia Less.
—
Arbores vel frutices, foliis alternis, spinoso-dentatis vel
intege rrimis. Capitula apicibus ramulorum, solitaria vel 2-
. sessilia. Inve — campanulatum vel turbinatum.
Flores 4—150, lutei vel aurantiaci, isomorphi, herma-
— corolla шы pentasecta vel profunde penta-
Ча. Antherae. appendicibus connectivalibus caudatis
w
ec
=
* abruptis, appendicibus basalibus integerrimis vel la—
ciniatis. Pappus uniseriatus vel biseriatus.
Shrubs or small trees, monoecious. Leaves alternate, pet-
iolate or shortly petiolate; obovate; oblong, or elliptic with
margins spinose-dentate
upper surface generally Hee lower
tomentose (with flagellate hairs). — homagamous,
scoid, sessile, solitary at the tip of the branches (rarely
2 or 3). Involucre campanulate or turbinate. Phyllaries in
1 to 10(15) series, dorsally tomentose. Florets 4 to 150,
isomorphic, tubulose,
yellow or orange, hermaphroditic,
)pendages
deeply to very deeply 5-lobed. Anthers with aj
caudate (rarely abruptly apiculate), and — smooth,
occasionally laciniate tails. Style bilobate or shortly bifid;
style brane hes rounded, dors ally glabrous. Cypsclas with
of the same length (rarely biseriate with a reduced number
of outer short bristles) all thin.
Twe nt y-six spec les: ©. attenuata, G. buchii, G. calcicola,
pha G.
par vifolia. G. — G. ріс
shaferi, С.
sagraeana, б. Lortuensis, G
Distribution. Domini-
Bahamas Islands, Cuba, Haiti,
can Republic, i
Puerto Rico
Observations. As pointed out by Jervis (1954) and Ca-
Volume 89, Number 4
2002
Freire et al.
Gochnatia Morphology
brera (1971), Bf рн D. Don (Trans. Linn. Soc. Lon-
don 16: 295. onsidered an unknown :
cause Its type ini ilicifolia D. Don
Joannes Fraser" 'imen in the Lambert herbarium) has
never been locate d. The original diagnosis of Anastraphia
does not agree with the later interpretation of the genus
6).
by de Candolle (1838: 26
0) is c
Gochnatia sect. Discoseris
O —
Endlicher) € cabrera, Revista
6: 7
—
NS
"
Ш)
%
A
^A
=
=
~
Z
c
c
-—
=
Post & Kuntze, Lex. Gen. Cima
1904, nom. rfl. TYPE: Seris discoidea Less. Di
Gochnatia discoidea (Less.) Cabrera)
supe
— monoecious. Leaves alternate, petiolate;
wate-elliptic with margins entire or denticulate;
— veined; glabrous or tomentose on both sides
i — homogamous. discoid.
solitary or arranged in scapose
pseudocorymbs. — те коо to turbinate. Phyl-
laries in 4 or
long-pedunculate and
orsally tomentose. Florets numer-
ous (ca. 30), yellow, сие , isomorphic, tubulose,
deeply 5-lobed. Anthers with appendages abruptly apic-
ulate and laciniate tails. Style — bifid; style branches
rounded, dorsally glabrous. Cypselas with duplex and
glandular hairs. Pappus uniseriate with all bristles of the
same length and thin
5 ser ies,
Three species: G. amplexifolia, G. discoidea (type spe-
cies), G. suffrutescens.
Distribution. Southeastern Brazil.
G n S. E. Freire, L. Katinas €
TYPE: Goc hnatia arborescens
— — ‘ted here).
Goc — sect.
S;
Arbores vel frutices, foliis — Mis iiie Capit
ula pauca, ibus nite od
glomerata. Involucrum лл Flores 12-50, lu-
tei, ison — bi maphroditi, corollis tubulosis, pentasec-
is. Antherae appendicibus connectivalibu
abruptis,
sessilia subsessi la,
s attenuatis vel
appendicibus basalibus integerrimis.
biseriatus vel uniseriatus.
Small trees or shrubs, monoecious.
shortly petiolate; ovate to elliptic with
Pappus
Leaves alternate,
margins entire or
dentic — pinnate Ny veined; upper surface glabrous or
tomentulose, lower ace usually der
AE Malla аў. Cda homogamous, discoid, ses-
sile, few together in terminal glome —
kG ory р. Inoue ‘re campanulate. Phyllaries in 5
6 series (rarely 8 to 10), dorsally tomentose or gla
Florets 12 to 20 bu ca. 50), yellow, ——
isomorphic, tubulose, deeply 5- lobo d. Anthers y
asely tomentose
ped
with ap-
pendages abruptly apiculate (rarely —— and smooth
tails. Style shortly bifid or bilobate; style branches round-
ed, dorsally glabrous. Cypselas with duplex and glandular
hairs. Pappus biseriate of numerous scabrid bristles, with
a reduced number of outer short bristles (rarely uniseriate
with all bristles of the same length), all thin
Three spec ies: G. arborescens (type species), G. magna.
G. ригриѕи.
Distribution. Northern Mexico.
Gochnatia Kunth sect. Gochnatia, Nov. Gen. Sp. 4: 15.
1818. TYPE: Gochnatia vernonioides Kunth.
Shrubs, monoecious. Leaves alternate, shortly petiolate;
ovate or ovate-elli with margins entire (rarely dentic-
ulate); pinnately, occa ia anal subtri-veined; upper surface
generally glabrous or ы: клы ower surface densely
tomentose (м flagellat Майа), Сари e
omentose above. Florets
pe in
series, — lly tomentose or only
to + ellow, ече a litic,
— 5- iud ers with appendages caudate, more
rare ly abruptly apic eae and iss tails. Style shortly
bifid; style branches rounded, dorsally glabrous. Cypse las
with la hairs m Mosis glandular hairs. Pappus bis-
eriate of numerous scabrid bristles s, with a
ber of outer, short bristles, all thir
tubulose,
reduced num-
Seven species: G. arequipensis, G. boliviana, G. carden-
asii, G. curv a ora, G. patazina, G. vargasti, G. vernonioides
(type specie
Distribution. Andes of Peru, Bolivia, and northwestern
Argentina.
Goc 'hnatia sect. е (Lessing) DC., Pre
7(1): 24. 1838. Gochnatia subg. He leia
PA Syn. а Compos.: 103. 1832. TYPE: Gochn-
atia cordata Less.
Gochnatia sect. Moquiniastrum Cabrera, Revista Mus. L
Plata 12, Secc. Bot. 66: 73. 1971. TYPE: А
polymorpha Less. (= G. polymorpha (Less.) Cabrera).
—
Shrubs or trees, commonly By nodioeci ‘ous or polvgamous
Leaves alternate, petiolate or shortly petiolate:
ovate, elliptic (rarely linear or cordate) with margins en-
tire, rarely denticulate; pinnately veined; upper surface
glabrous (rare ly tomentose), lower surface densely tomen-
with 2-, 3- to 5-armed hairs, occasionally flage llate)
ири heterogamous (homogamous), dise ¡form —
discoid), subsessi
dioecious.
tose (
ssile to pedunculate, numerous arranged in
terminal and usually loose leafy pseudopanicles. Involucre
oblong to campanulate. — in 3 or 4 series, dorsally
tomentose. Florets 9 to 15(50), creamy or white, tubulose,
deeply 5-lobed, subdindtiphie (occasionally isomorphic):
functionally female, corollas slightly zygomorphic,
straight lobes: — ac —
Anthers кү ар
and laciniate or sm iUis ails. Style s
bilobate): style bra
selas — duplex. epe nl and f
biseri of numerous scabri
with
with resu-
pinate lobes.
ж
d а
метав of outer short bristles, all thin, and the longest
are plumose at the ay
ea, G. bar-
cies), G. den-
Twe 2n -one speci
rosii,
: G. argentina, G. argyr
‚ blanc hetiana. G. cordata (type spec
on G. Дог
ш. G. — iniana, G. mollissima, G. olig
phala, G. orbiculata, G. Dan c. аиа G. "e
chra, G. ramboi, G. rusbyana, G. sordida, G. velutina.
OCE=
Distribution. Andes of Peru and Bolivia, eastern Bra-
‚ Paraguay, Uruguay, and central-eastern Argentina
N
Сос hoatia sect. Leucomeris (D. Don) Cabrera, Revista
Mu
. La Plata 12, Secc. 66: 128. 1971. Leucomeris
D. 3o Prodr. Fl. Nepal.: 169. 1825. Gochnatia
ru Less., Syn. Gen. SURE
. Don)
sochnatia spectabilis (D
Less. (= — —— spectabilis D. Don).
550
Annals of the
Missouri Botanical Garden
Trees or shrubs, monoecious. Leaves alternate, shortly
pe tiolate; elliptic, margins entire or denticulate; pinnately
— ined; Md r surface generally parous to tomentose,
surface densely tomentose (with flagellate > hairs) o
=
تب
tending to the peduncle. Florets 4
(rarely yellow), hermaphroditic, isomorphic, tubulose,
deeply 5-lobed. Anthers with appendages attenuate (rarely
abruptly apiculate) and lac iniate tails (rarely smooth).
I
ily with glandular hairs. Pappus Бегай of nu-
merous scabrid bristles, with a reduced number of outer
short bristles, all thin
Five species: G. hypoleuca, б. RON. G. smithii, G.
decora, C spectabilis (type species
Distribution. Mexico, Brazil, + region of Bolivi:
and Argentina, and southeastern Asi
Gochnatia sect. Pentaphorus (D. Don) DC., Prodr. A 1):
24. 1838. e D. Don, Trans. Linn. S
16: 296. 1830. subg. P uei
). Don) Hook. & Arn., Cor Bot. Mag. 1: 108.
Ta a PE: * — "foliolosa D. Don ex Hook.
& Ап
—
=
monoecious. Leaves alternate, sessile; linear-
ate with margins entire or — por⸗
and low-
Shrubs,
ovate or linear-obov:
tion denticulate; pinnately or three-veined: upp
er surfaces glandulate (with or without ode hairs).
Capitula homogamous, discoid, sessile or subsessile, nu-
erous in leafy glomerulose pseudorac Involucr
1 cemes. re
campanulate. Phyllaries i in 4 to 6 series, dorsally glabrous
or — with ciliolate margin. Florets 5 (rarely 3
r 10 to 20), white or lilac, hermaphroditic, isomorphic.
lee deeply 5-lobed. Anthers with appendages
abruptly apiculate md smooth or laciniate a pare bi-
lobate or shortly bifi d; style branches rounded, dorsally
relatively vids and long. and the others short and thin.
Two species: G. foliolosa, G. glutinosa (type species).
Distribution. Western Argentina and central Chile.
Gochnatia sect. Rotundifolia 5 . E. Freire, L. Kat
Sancho, sect. nov. TYPE: Gochnatia rotundifolia
Less. (selected here).
Frutices, foliis alternis, integerrimis vel denticulatis.
Capitula apicibus ramulorum solitaria, sessilia. Involu-
crum campanulatum; bracteis involucralibus subglabris,
margine ciliatis. Flores multi, E — herma-
phroditi, corollis tubulosis, pentasectis. Antherae appen-
dicibus connectivalibus attenuatis р appendici-
bus basalibus laciniatis. Pappus biseriatus, setosus.
Shrubs, monoec ious. Leaves alternate, shortly petiolate;
broadly ellip ulate; three
veined, ru see on both surface PUDE
leaves with flagella te hairs). Cia homogamous, dis-
Phyllaries
c with margins entire to dent
2
ur
z
=>
5
t3
lose, deeply 5- ва е
and laciniate tails. Е pour style branches rounded,
dorsally glabrous. Cypselas with duplex and glandular
hairs. Pappus biseriate of numerous scabrid bristles, with
a reduced number of outer, short bristles, all thin
One species: G. rotundifolia.
Distribution. Southeastern Brazil.
EL GENERO QUERCUS
(FAGACEAE) EN EL ESTADO
DE MEXICO!
Silvia Romero Rangel, Ezequiel Carlos
Rojas Zenteno? y María de Lourdes
Aguilar Enríquez
RESUMEN
Con base en trabajo de campo y E herbario, se re — en 23 especies de Querc us para el Estado de México. Diez
pertenecen a la sección Quercus y a la sección
— distribuidas ше Centro 8 ıérica y sólo Quercus rugosa se encuentra
. El trabajo incluye una clave dicotómica y des
os Estados Unidos
Lobatae. Quince de las especies son endémicas de M
éxico, siete
en el norte de la frontera mexicana. en
‘ripciones morfológicas de cada uno de los taxa.
re ней nte, para cada especie se proporcionan datos e nir ree os, fenológicos y ecológicos
ABSTRACT
Based on 1 field and herbarium work, in this paper 23 species of Quercus are rec ognized for the State of México. Ten
xtend to Central America, and only Quere
1 Quercus = 13 to section Lobatae. Fifteen of the species are endemic to Mexico, seven
is rugosa is found north of the Mexican border in the United States.
The
paper includes a dichotomous key and эйе ыг ‘al descriptions for all the taxa, as well as ethnobotanical, pheno-
ogical, and ec al data
"Kay words: Fagaceae, — Quercus.
El género Quercus se distribuye mundialmente
en zonas templadas y subtropicales del hemisferio
norte. Se calcula que está conformado aproxima-
damente por 500 especies distribuidas mundial-
mente. México es el país que posee el mayor nú-
150
distribuidas en las zonas monta-
mero de especies del mundo, entre 135 y
(Nixon, 1993b),
ñosas de todos los estados y territorios, a excepción
del sur en Yucatán y Quintana Roo (Rzedowski.
1978
las especies del género Quercus se les en-
cuentra formando bosques de encino, comunidades
muy características de las zonas montafiosas de
México. De hecho junto con los miembros del gé-
nero Pinus constituyen la mayor parte de la cu-
bierta vegetal de áreas de clima templado y semi-
húmedo; sin embargo no se limitan a estas
condiciones, pues penetran en regiones de clima
caliente formando también bosques, no faltan en
las francamente húmedas, siendo elementos del
bosque tropical perennifolio y bosque mesófilo de
montaña, y aún existen en las semiáridas siendo
parte del matorral xerófilo, asumiendo con frecuen-
cia forma arbustiva (Rzedowski, 1978).
Existen dos trabajos que intentan registrar las
especies de la entidad en estudio. Uno es de Mar-
tínez (1954) titulado “Encinos del Estado de Méx-
ico”: en él se refieren 45 especies, número que se
ha modificado por la gran cantidad de sinónimos.
El otro es de Espinosa (Rzedowski & Rzedowski.
1979), quien publicó, dentro de la “Flora fanero-
eámica del Valle de México”, los encinos de esta
zona. Una porción de esta región pertenece al Es-
tado de México; en él se mencionan 13 especies
para el área de estudio. Otros trabajos importantes
que consideran especies de la entidad estudiada
son los de Muller y MeVaugh (1972), MeVaugh
(1974) y de Aguilar y Romero (1995).
El Estado de México ocupa una superficie de
22.500 km? y está comprendido en dos provincias
fisiograficas: la del Eje Neovolcánico que abarca la
mayor parte de la superficie del norte, con altitudes
! Se agradece a los encargados de los herbarios CHAPA, CODAGEM, ENCB, INIF, IZTA y MEXU por el préstamo
A los Doctores Jerzy Rzedowski R.,
de sus colecciones.
Patricia Dávila
A. y Rafael Lir
a Š. por sus comentarios y
valiosas sugerencias. A la Bióloga Adriana Bernal P. por la realización de los dibujos de ә ejemp
= comunicación pertenece al proyecto encinos de México que se lleva a cabo en la UBIPRO de lat UNAM Campus
Iztacala.
2? UBIPRO. _ NAM Campus lztacala. Av.
р 4090. sromero@servidor.unam.t
5 Педа IZTA-UN. AM Cam
México. C.P. 54090. zenteno@servidor.unam.
mpus Iztacala. nh De los Barrios s.n
De ы Barrios s.n. Los Reyes Iztacala Tlalnepantla, Estado de México,
. Los Reyes Iztacala Tlalnepantla, Estado de México,
ANN. Missourt Bor. GARD. 89: 551—593. 2002.
Annals of the
Missouri Botanical Garden
que van de 2000 a 5452 m y la de la Sierra Madre
del Sur que abarca el extremo SE, donde las alti-
tudes son de 600 a 3000 m (INEGI, 1987)
Los encinos son abundantes en las regiones mon-
tafiosas de la entidad estudiada, formando parte de
los bosques de coníferas, bosques de Quercus,
bosques mixtos de Pinus—Quercus y Cupressus—
Quercus; aunque también habitan en otras comu-
nidades tales como bosque mesófilo de montaña,
bosque tropical caducifolio, no faltando en matorral
xerófilo, bosque de galería y pastizal.
Quercus es considerado como un género taxo-
nómicamente complicado, debido a su gran varia-
bilidad morfológica, a la capacidad de formar hí-
bridos y a que sus estructuras reproductivas no se
han estudiado lo suficiente (Romero et al., 2000).
Esto ha llevado a la existencia de una gran canti-
dad de sinónimos.
MATERIALES Y MÉTODOS
El trabajo se basó fundamentalmente en la re-
visión y estudio de los ejemplares existentes en los
herbarios CHAPA, CODAGEM, ENCB, INIF, IZTA
y MEXU.
Se obtuvo la variación morfológica regional de
los ejemplares citados y de las observaciones rea-
lizadas en campo por los autores.
Para determinar la aplicación de los nombres se
revisaron las descripciones originales y los ejem-
plares tipo provenientes de herbarios; cuando estos
últimos no estuvieron disponibles para su análisis,
se utilizaron en su lugar las fotografías de los tipos
contenidas en Trelease (1924). Estos últimos y los
herbarios donde están depositados se citan como lo
hace dicho autor. En este trabajo se incluyen sólo
los sinónimos encontrados para el Estado de Méx-
ico.
Las descripciones morfológicas se realizaron de
acuerdo al formato del tratamiento del género Quer-
cus para la Flora Novo-Galiciana (Mc Vaugh, 1974)
El hábitat, las descripciones morfológicas, la dis-
tribución geográfica y los dibujos de las especies
están basados en los ejemplares revisados para
cada especie.
TRATAMIENTO TAXONÓMICO PARA EL GÉNERO
QUERCUS EN EL ESTADO DE MEXICO
Son 23 las especies reconocidas en este trabajo;
10 pertenecen a la sección Quercus y 13 a la sec-
ción Lobatae, de acuerdo a la clasificación del gé-
nero Quercus propuesta por Nixon (1993a).
TYPE:
Quercus L., Syst. Pl. ed. 2, II, 994. 1753.
(fide ING)
1938).]
Quercus robur L. . [Sinonimia com-
pleta en Camus
Árboles o arbustos;
reunidas en las puntas de las ramas; estípulas su-
monoicos; yemas foliares
buladas o liguladas, generalmente decíduas, a ve-
ces persistentes, más bien asociadas con las yemas
que con las hojas; hojas alternas, generalmente pe-
cioladas, nunca totalmente sésiles; amentos mas-
culinos largos y colgantes, flores con el cáliz 5-
lobulado, fusionado a un perianto que envuelve a
los estambres; estambres de 5 a 10, libres, con an-
teras cortas y filamentos delgados; flores femeninas
en forma de racimo reducido con un raquis leñoso
corto o largo y con una o varias flores, el cáliz con
6 lóbulos que se adhieren a la base de los estilos
y se fusionan en un tubo; pistilo de 3 carpelos que
forman un ovario trilocular, cada lóculo con 2 óvu-
los; estilos 3, libres; fruto unilocular con una se-
milla, los otros 5 óvulos son abortados; semilla en-
vuelta en una cubierta rígida formando la bellota
que está protegida parc ‘ialmente en su base por una
cúpula cubierta de escamas; N = 12
I. Quercus sect. Quercus
Base del perianto femenino adnado al ovario; es-
tilos cortos y anchos; pared interna del endocarpo
glabra; escamas de la cúpula aquilladas, engrosa-
das en la base; dientes de las hojas mucronados;
óvulos abortivos basales.
1. О. deserticola Trel.
2. Q. frutex Trel.
3. О. glabrescens Benth.
4. Q. glaucoides M. Mart. & Gal.
5. Q. laeta Liebm.
6. Q. magnolufolia Née
7. Q. obtusata Humb. & Bonpl.
8. Q. peduncularis Née
9. Q. rugosa Née
10. Q. splendens Née
П. Quercus sect. Lobatae Loudon., Hort. Brit.
385. 1830. TYPE: Quercus aquatica Walt.
(lectotipo, designado por Nixon (1993a)).
Base del perianto femenino libre; estilos por lo
general alargados, lineares; pared interna del pe-
ricarpo pubescente; escamas de la cápula, no aqui-
lladas, no engrosadas en la base; dientes de las
hojas, por lo general aristados; óvulos abortivos en
posición apical o lateral.
Q. acutifolia Née
12. Q. candicans Née
13. Q. castanea Née
Volume 89, Number 4 Romero et al. 553
2002 Quercus en el Estado de México
14. Q. conspersa Benth. rarlas es que Q. frutex posee pubescencia en el
15. Q. crassifolia Humb. & Bonpl. envés de las hojas que permite ver la epidermis,
l6. Q. crassipes Humb. & Bonpl. formada por tricomas estrellado-estipitados con me-
V7. Q. dysophylla Benth. nos de nueve rayos; mientras que Q. repanda posee
18. Q. elliptica Née pubescencia muy densa que no deja ver la epider-
19. Q. hintonii E. F. Warb. mis, formada por tricomas estrellados sésiles o cor-
20. Q. laurina Humb. & Bonpl. tamente estipitados con más de 15 rayos.
21. Q. mexicana Humb. & Bonpl.
22. Q. scytophylla Liebm. Il. QUERCUS SECT. LOBATAE
23. Q. urbanii Trel. Quercus affinis. Schiede. Citada por Martínez
De los 45 nombres citados en la bibliografía para (1954), ips à similar a Quercus laurina, con la que
el Estado de México, 20 resultaron ser sinónimos {Че confundida. La pU distingue por ——
de las especies tratadas en este trabajo y las cinco Ye™mas foliares de forma conoidal, base de las hojas
siguientes las hemos considerado no existentes en cuneada, nervaduras planas, y distribución en alti-
la entidad estudiada; se indican las secciones a las tudes menores a los 2400 = mientras que Q. lau-
que pertenecen (Nixon, 1993a). rina posee las yemas ovoides, la base de la hoja
atenuada о redondeada, nervaduras realzadas y dis-
1. QUERCUS SECT. QUERCUS tribución en altitudes mayores a los 2200 m.
Quercus мы Née fue confundida también
Quercus sanchez-colinii Mart. Citada por Martí- por Martínez (1954) con Q. laurina. La primera
nez (1954), seguramente corresponde a una varia- presenta de 15 a 25 а primarias еп las
ción de Q. laeta debido a la similitud morfológica hojas y se distribuye en la vertiente del Pacífico,
entre ellas. Sin embargo no se han vuelto a colectar en altitudes de 600-2000 m. La segunda posee un
4 a 12); se
—
ejemplares con hojas de borde ondulado como las número menor de nervaduras primarias
observadas en el tipo de la primera especie. le encuentra en estados de la vertiente del Pacífico
Quercus repanda Humb. € Bonpl. Citada por Es- y del centro del país, en altitudes de 2240-3150 m.
pinosa (Rzedowski & Rzedowski, 1979), fue con- Quercus aristata Hook & Arn. Citada por Mar-
fundida con Q. frutex debido a que ambas presen- — tínez (1954), no se encontró en los herbarios revi-
tan formas arbustivas y similitud en la morfología sados, ni se logró colectar en los sitios referidos en
de la hoja. Una diferencia que hace posible sepa- la descripción original.
CLAVE DE LAS ESPECIES DE QUERCUS DEL ESTADO DE MEXICO, MEXICO
la. rue de borde entero, algunas veces mucronadas o aristadas en el ápice, pero sin dientes laterales
nvés de las hojas glabro o con tricomas concentrados en las axilas de las nervaduras o a lo Em de
las mismas, frecuentemente glandular.
Hojas comunmente glauca is, ápice nunca aristado; cúpula con escamas engrosadas en la base;
pared interna del pericarpo glabra
la. Superficies maduras labras | por — to o con algunos tricomas cerca de la Pu nerva-
duras rojizo amarillentas 0000 ). glaucoides
4b. Superficies maduras con tricomas a lo largo de la nervadura central, NE. ve des |
NI ORES AT 10. О. —
3b. Hojas no glaucas, ápice comunmente aristado; escamas de la cüpula no engrosadas en la base
sa.
Ramillas y pecíolos — pubescentes —— Be Ж ыны 18. Q. elliptica
5b. стт y pecíolos glabrescentes o glabros.
jas de 11-17 em de largo: con el envés con abundantes tricomas glandulares .........
RR И КЕЛЕНКЕ IO EN. A AA — 14. Q. conspersa
6b. Hojas de 5-11 cm de largo; con escasos tricomas glandulares |... 20. Q. —
2b. Envés de la hoja pubescente, con los tricomas uniformemente distribuidos, — la lámina, no con-
centrados en las axilas de las nervaduras o a lo largo de las mismas
7a. Apice de la hoja aristado; pared interna del pericarpo lanos:
Ва. Envés de la hoja con tricomas contortos, epidermis — RE E XN 2]. Q. mexicana
8b. Envés de la hoja con tricomas no contortos; epidermis ——
9a. Pubescencia amarilla 17. Q. dysophylla
9b. Аган аду ө grisácea.
10 Narvadurda conspicuamente elevadas en el envés, de 5-12 en cada lado, bellota
de 5-15 mm de largo . castanea
10b. Nervaduras ligeramente elevadas en el envés, de 10-19 en cada TM bello ta
de 12-30 mm de largo 16. Q. crassipes
+
554 Annals of the
Missouri Botanical Garden
Tb. Apice de la hoja con un mucrón; pared interna del pericarpo glabre
11а. Ar ‘bustos poe de 0.4 poa m de alto; n comunmente de 2—4.5 cm de —
pula del fruto de 7-13 mm diár — ). frutex
11Ь. Шы 's — de 2— ij m de а. һојав comunmente de 4—7.5 ст de largo: cpu de
fruto de 14-17 mm diár — 1. Q. deserticola
Ib. Hojas con el borde dentado, ondulado o con n los bordes aristados o mucronados.
12a. Hojas con el borde ondulado o con dientes mucronados, nunca aristado.
13a. Envés de la hoja glabro, glabrescente o con tricomas concentrados en las axilas de las nervaduras
o a lo largo de las mismas.
14a. Hojas con el envés glauco.
15a. Superficies — — por completo; árboles de 4-10 m de alto; My de 4—
4.
>
SIAM de dina ше oides
I5b. Superficies maduras con tricomas que tienden a concentrarse en la nervadura с * мга
árboles de 5 m de alto; estípulas de 5-8 mm de largo . Е Wu ). s andén
14b. Hojas con el envés no — е 3. —
). g
13b. Envés de la hoja pubescente, con los tricomas uniformemente distribuidos sobre la TEA no
concentrados en las axilas de las ne — o a lo largo de las mismas.
16a. iue densamente pubescente
Arbustos rizomatosos de 040-2. 50 m “ alto; hojas comunmente de 2—4.5 cm de
argo; ctipula del fruto de 7-13 mm dia 2. Q. frutex
Vib. н 's pequefios de 2-7 m de alto; hojas comunmente de 4—7.5 cm de largo; pis
Q
-]
a
—
el fruto de 14—17 mm diám . deserticola
16b. Ramillas labras o glabrescentes.
Tricomas del envés de las hojas estipitados
19a. Envés de las hojas con indumento sinclar frutos sésiles о en ip culos de
nm de largo... s l. Q. dy —
19b. Envés de las hojas con indumento o blanquecir ino; . frutos en n pedúne a de 3.5-
oni ае largo ers ырш e ыманы Мема ымы ШЫ 8. Q. —
J rg
18b. P — 'encia del envés formada ү por tricomas sésiles o con un estípite Pi
20a. Tri
6. Q. magnoliifolia
=
4
ricomas sésiles 0
20b. Tricomas con un A DB A corlo.
2la. Epidermis del envés de las hojas con algunos tricomas puse
AD MA A e NE E ы а айлына E Ыссар ан аат нышын елаш 0. laeta
21b. — rmis del envés de las hojas con abundantes tricomas glandt is
— de la hoja con mucrones de hasta 2 mm de largo; bellota
vide o angostamente elíptica — 9. O. rugosa
22b. Marge n de la hoja con mucrones cortos, menores de 1 mm, que se
'urvan hacia el envés; bellota globosa o cilíndrico-ovoide |...
7. Q. obtusata
12b. Hojas con borde aristado.
23a. Envés de la hoja glabro, o con tricomas concentrados en las axilas de las nervaduras o a lo largo
de las mismas.
24a. Epidermis — yemas de 1.5—4 mm de largo; árboles altos, de hasta 30 m, crecen en
10-315 20.
altitudes de AAA AAA Q. laurina
24b. Epidermis no — sa; yemas ed mm de [ox árboles Шош; de hasta 12 m de alto,
‹ altitudes menores de 2 y | 11. Q. acutifolia
23b. Envés de la hoja con los tricomas тё memente distribuidos sobre la lámina, no concentrados
1 las axilas de las nervaduras o a lo largo de las mismas.
25a. Envés de las hojas blanquecino, — m она рог tricomas sésiles
26a. Hojas con el haz verde lustroso, de 1 nervaduras en cada lado; yemas de 3-5 mm
de — estípulas de 10-15 mm гу ie frutos de 20 mm de largo y de 17 mm de
AAA 12. Q. candicans
206b. Hojas con el haz ale grisdceo, no Шанбай, de 5-9 nervaduras en cada lado; pec "=
de 14 cm de largo; — de hasta 5 mm de ا frutos de 8-10 mm de larg
de 10 mm de ancho — s 22. Q. : —
25b. Envés ids las hojas no blanqueci ino, indumento formado por tricomas estipitac ados
27a. Envés de las hojas con ose grisáceo; hojas elfpticas, elíptico- oblonga, oblan-
ceoladas o lanceoladas |... s . castanea
27b. Envés de las a con indumento amarillento: hojas obovadas, oblongo- ola
suborbiculares u orbiculares, ovado-elípticas, raramente elípticas.
28a. Ramillas de 5-11 mm diám., — panduriformes, suborbic — orbiculares,
obovado-elípticas u ovadas, de 15-30 em de largo y de 17-34 em de e :ho
. M e cias 3. Q. urbanii
28b. — de 2-5 mm diám., hojas planas, lane eoladas, ovado lanc coladas obo-
adas, oblongo-obovadas o elípticas, de 4-20 em de largo y de 3-12 em de
ancho.
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
29a. Epidermis de la hoja ampulosa; cúpula del fruto hemisférica; — ovoide
15. Q. crassifolia
29b. Epidermis de la hoja lisa; cúpula del fruto poc -uliforme a | pate Ew. be-
—
—
1. Quercus e Trelease, Mem. Nal
Acad. Sci. 20: 79, pl. 113. 1924. TIPO: Méx-
ico. p. Uhde 309 (B).
Quercus alveolata Trel., Mem. Natl. Acad. Sci. 20: 80, pl.
114. 1924. TIPO: Cerro del Gavilán, Puebla, Purpus
1 (NY).
em. Natl. Acad. Sci. 20: 81, pl.
Quercus texcocana Trel.,
117. 1924. TIPO: e rro Texcotsingo, cerca de Te "x
coco, Endlich 653 (B).
Arbolillo de 2—7 m de alto, corteza gris; ramillas
de 1-3 mm diám., densamente pubescentes, verde-
amarillentas, el indumento se ennegrece con el
tiempo, formado por tricomas estrellados: lenticelas
de hasta 1 mm, blancas, visibles sólo en ramillas
viejas; yemas de 2-3.5 mm de largo, ovoides, es-
camas escariosas con los bordes ciliados; estípulas
de 3-5 mm de largo, lineares o filiformes, pilosas,
rojizas, persistentes en hojas Jóvenes y yemas; ho-
jas jóvenes rojizas, haz verde, con el indumento
formado por abundantes tricomas estrellados cortos.
estipitados, envés más pálidos, con pubescencia
más densa formada también por tricomas estrella-
dos estipitados, pero con las ramas más largas: ho-
jas maduras oblongas, elfpticas, elfptico-oblongas u
obovadas. lámina (2.5-)4-7.5(-8.5)
X (1.3—)2—3.5(—4.5) ст, ápice agudo u obtuso, mu-
subcoriáceas,
cronado, base cordada o subcordada, borde entero.
revoluto, ondulado o dentado, 2 a 5 dientes mucro-
nados de cada lado, frecuentemente asimétricos;
nervaduras de 6 a 9 de cada lado, que se continúan
en el diente; haz verde lustroso con tricomas estre-
llados cortos, estipitados, uniformemente distribuí-
dos, más abundantes en la nervadura central, ner-
vaduras impresas; envés pálido, con abundantes
tricomas estrellados, estipitados, con las ramas más
largas que los del haz y tricomas glandulares rojizos
sobre la epidermis ampulosa y papilosa, nervaduras
elevadas; pecíolos de 2-5
] mm diám., pubescentes al igual que las ramillas,
mm de largo, de 0.5—
base engrosada; amentos femeninos con 3 a 10 flo-
res en pedtinculos de hasta 30 mm, densamente
pubescentes: fruto anual, solitario o en grupos de
2 о З sobre pedünculos de 2-9 mm de largo: cúpula
hemisférica, de 14—17(—20) mm diám.,
con pubescencia blanca, las de la base engrosadas,
las escamas
las superiores menos pubescentes, ápices obtusos,
ligeramente elevados; bellota ovoide, pared interna
del pericarpo glabra, de 11-13(219) mm de largo,
de 11-15 mm diám., un tercio de su largo incluida
en la cúpula. Figura
ota globosa a comprimida
J. Q. hintonii
Reconocimiento. Quercus deserticola se reco-
noce por ser un arbolito con las ramillas densa-
mente pubescentes y hojas de hasta 7.5 em de largo
con bordes revolutos y sin aristas.
Distribución y hábitat. En México en los esta-
os de Distrito Federal, Guanajuato, Hidalgo, Ja-
lisco, Estado de México, Michoacán y Querétaro.
En bosque de Quercus, pastizal y matorral xerófilo,
asocia con Alnus
se
2600-2800 m.
Fenología.
y Cupressus, en altitudes de
Florece en abril y fructifica de julio
a diciembre.
Nombres populares y usos. Encino, encino to-
cuz, encino chico.
Su corteza se usa para curar encías, amacizar los
dientes y en curtiduría; su madera como leña, para
fabricar pulpa para papel, carbón, postes para cer-
ca, arados, cabos para herramientas y horcones
(González. 1986: Bello € Labat. 1987).
MÉXICO. Estado de Méxi-
co: Seule de San Luis Aculco, Capse 99
(ENCB "Асар. Ciudad Adolfo López Mateos y cerca-
nías, Маны 3 28314 (MEXU); parte alta del Cerro Chi-
— — de Ciudad Adolfo López Mateos, Rzedowski
Ejemplares examinadas
¿spinos
32588 (ENCB); Coyotepec, Sierra de Alcaparrosa, Fer-
n sa s.n. (IZTA); Sierra de Alcaparrosa, Callardo 1 B
(ТИТА): Sierra de — Romero 153 (ИТА); Eca-
tepec de Morelos, 6 km al E de San Cristóbal Ecatepec.
Rzedowski 32154 (ENCB); Huixquilucan, cerca de la Pre-
sa El Capulín, Fraccionamiento La Herradura, Rzedowski
25875 (ENCB); Dos Ríos, Rzedowski 35586 (ENCB); Ix-
— a, parte alta del Cerro del Wen Rzedowski 29655
NCB); Teoloyucan, Sierra de rosa, Núñez 1718
(ZT, "s Te potzotlán, Teporzotlán, Mia 30 (ЕМСВ); Si-
erra 06 Alcaparrosa, Núñez ZTA); parte baja de la
ı de Alcaparrosa, 2 km al NN V A oo Rze-
e эи shi 29912 (ENCB); parte alta de la Sierra de Alca-
rosa, cerca de la estación de * mont: Re
59910 (ENCB): Texcoco, Cerro de '
E de Texcoco, Pulido 305 (C HAPA.
bón. Cañón E de la Presa Taxhimay, Rojas & Romero 3315
(IZTA).
2. pos reus frutex Trelease, Mem. Natl. Acad.
20: 82, pl. 120. ee TIPO: México.
ени Bourgeau 68
Arbusto rizomatoso de 0.40—2.50 m de alto, a
veces es un árbol de hasta 7 m de alto; corteza
conformada por escamas cuadrangulares, de color
gris opaco; ramillas de 1—1.5 mm diám., densa-
mente pubescentes, pubescencia persistente, for-
556
Annals of the
Missouri Botanical Garden
` AY
j ys
Y Ay oue 0
ў 6
SEE
SAN vie
е :
0.6mm
Figura 1. Quercus deserticola. —A. Rama. —B. Fruto. —C. Tricomas. —D. Rama. (A-C: Núñez 1663; D: Camacho
379.)
mada por pelos estrellados, de color amarillento a
grisáceo; lenticelas blancas, de hasta 1 mm de lar-
go, a veces protuberantes y entonces muy evidentes
a través de la pubescencia, yemas esféricas a ovoi-
des, de 1-3 mm de largo, con escamas pilosas en
los bordes; estípulas lineares, de 3—5 mm de largo,
pilosas principalmente en la base y en el ápice,
glabrescentes con la edad, con frecuencia en las
hojas maduras; hojas maduras subcoriáceas, elíp-
tico-oblongas, ovado-lanceoladas u oblanceoladas,
lámina (1.5-)2 4.56) X (0.5-)1-2(-2.9) ст, ápi-
ce redondeado o agudo, mucronado, base redon-
deada o subcordada, borde entero, ondulado o con
2 a 4 dientes en las % partes superiores, revoluto;
nervaduras primarias de 6 a 11 en cada lado; haz
lustroso, con pelos estrellados dispersos, abundan-
tes en la nervadura central cerca del pecfolo, ner-
vaduras impresas; envés amarillento, pubescente,
indumento formado por pelos estrellados estipita-
dos, con rayos largos extendidos que dejan ver la
epidermis ligeramente ampulosa y densamente pi-
losa; nervaduras elevadas: pecíolos de 1—4 X 0.5-
| mm, pubescentes; amentos masculinos de hasta
1.5 em de largo, raquis densamente pubescentes;
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
Figura 2.
perianto sésil, de 2 mm diám., lóbulos ciliados; an-
teras oblongas de 1 mm de largo, filamentos de 1.5
mm de largo: fruto solitario o en pares, sobre un
pedúnculo de 3-10 mm de largo: cúpula hemisfé-
rica de 7-13 mm diám., con escamas triangulares,
pubescentes, excepto en el dorso superior, engro-
sadas en la base; bellota ovoide, pared interna del
pericarpo glabra, de 5-11 mm diám., incluída %4
de su largo en la cúpula. Figura 2.
Reconocimiento. Quercus frutex se reconoce por
ser un arbusto rizomatoso, con ramillas densamente
pubescentes y hojas pequeñas con el ápice mucro-
nado y bordes sin aristas. Puede confundirse con
Q. microphylla; por ello es que otros autores la han
considerado existente en el Estado de México. Se
distinguen porque Q. microphylla posee el envés
de las hojas con pelos estrellados sésiles y Q. frutex
los presenta estipitados.
Distribución y hábitat. En México en los esta-
dos de Jalisco, Hidalgo, Estado de México, Mi-
choacán y Tlaxcala. Se le encuentra en bosques de
Quercus, pastizal y formando manchones densos en
matorral xerófilo y pastizal. Se asocia con Pinus,
Juniperus, Cupressus y Alnus; también se le en-
cuentra en vegetación perturbada, en altitudes de
2360—3000 m.
x MEAS ADT LD
RIYA
Y PES
ELE
KSSS
Quercus frutex.—A. Rama. —B. Fruto. —C. Tricomas. (Román 622.)
Fenología. Florece en junio y fructifica de julio
a septiembre.
Nombres populares y usos. Encino, encino com-
pasillo, encino chaparro.
Bello y Labat (1987) mencionan que se utiliza
para postes de cercas y como leña.
Ejemplares examinados. MÉXICO. Estado de Méxi-
сө: Atizapán, 3 km al NW de Ciudad Adolfo López Ma-
teos, Cruz 620 (ENCB); 4 km al N de Ciudad Adolfo Ló-
‚ López 19 (ENCB); Ciudad Adolfo López
Rzedowski 30776 (ENCB):
elos, Sierra de Guadalupe. al N de Coa-
epec, Rzedowski 15715 (ENCB); Coyotepec, Sierra de Al-
caparrosa, Gallardo 3, 4 (IZTA); Chalco, Tlachavote, A.
Ventura 3016 (ENCB); El Oro de Hidalgo, Bassoco, Rojas
& Romero 3327 (УТА); Huehuetoca, ladera Suroeste del
cerro Sincoque, Rojas & Romero 479 (IZTA); ladera E del
cerro Sincoque, Rojas & Romero 700, 701, 1623 (ПУТА):
Huixquilucan. alrededores de Dos Ríos, Román 326
ENCB): Fraccionamiento La Herradura, Rzedowski 25143
(ENCB): Ixtapaluca, 2 km al E de la Colonia Agrícola
Ávila Camacho, Rzedowski 37294 (ENCB); Naucalpan de
Juárez, Villa Alpina, Rzedowski 32554 (ENCB); Otumba
de Gómez Farías, Cerro Cuixi, Santa Bárbara, Rzedowski
16884 (ENCB); Teotihuacán, Cerro Patlachique, Chavelas
ES-1796 (ENCB); San Martín de las Pirámides, ladera N
de Cerro Gordo, Castilla & Tejero 388 (IZTA): Tenango de
Tepopula, Rancho San Luis Aculco, Hinton s.n. (ENCB):
Tepetlaoxtoc, La Venta, Ventura 678 (ENCB): Tepotzotlán.
Pp
vt
m |
a Ж D
=
S
558
Annals of the
Missouri Botanical Garden
®
Sierra de la Muerta (Sierra de Alcaparrosa), al N
Tepotzotlán, Lot & Wendt 143 (CODAGEM): — bon
del Cerro de la Cruz, dowski 33270 (ENCB); Cerro de
la Cruz, 6 km al — Rzedowski 37055
(ENC B); Texcoco, 20 km al NE de Texcoco, Cruz 1845
; ; 17 km al E de Texcoco, Charles & Janice e
3397 (С НАРА); 15 km al E de — :0, Muller
CB); Cerro a — 7 km al E de Texcoco, p
: 466 (ENCB, MEXU); San — Ixayoc, Ventura
845 (ENCB); Ville 1 on Carbón, Cañón E bs la Presa '
himay, Rojas & Romero 3312 (IZTA):
mero, S de la Presa del Consuelo, La Coln
28 (IZ TA); 3 km al N de Magú, Rzedowski 16850
2 km al NW de Cahuacán, Azedowski 3:
Zumpango, T km al N de Zumpango, Hernández s.n.
(ENCB)
3. Quercus glabrescens Bentham, Pl. Hartw. 56:
348. 1840. TIPO: México. Hidalgo: Hartweg
428 (В).
Arbol de 10—15 m de alto; ramillas de 1-2.5 mm
diám. con lenticelas de color claro, de hasta 1 mm;
yemas ovoides de 2.5-3 mm de largo, escamas ova-
das con margen ciliado; estípulas oblanceoladas, de
7 mm de largo, glabrescentes, escariosas, persis-
tentes en las yemas; hojas maduras subcoriáceas a
coriáceas, oblanceoladas, elípticas u obovadas,
lámina (3.5-)4-11 X (1.2-)1.7—4(—5.3) em, ápice
agudo, base redondeada, borde revoluto, cartilagi-
noso, ondulado o con 3 a 5 dientes mucronados a
cada lado en la mitad superior de la hoja, mucrones
de hasta 1 mm de largo; nervaduras primarias de
10 a 14 en cada lado, ascendentes, casi rectas, las
superiores pasan directamente a formar los mucro-
nes; haz de color verde obscuro contrastando con
el envés verde más claro, glabrescente, con pelos
glandulares simples sobre las nervaduras y pelos
estrellados sésiles, nervaduras impresas; envés gla-
bro o con pelos estrellados sésiles o con un estípite
muy corto, ásperos, sobre la nervadura principal,
epidermis lustrosa, lisa a papilosa, a veces ligera-
mente ampulosa, nervaduras elevadas; pecíolos de
3-10 mm de largo, de 0.5-1 mm diám.,
dumento formado por pelos estrellados sésiles y ás-
con in-
peros; amentos masculinos de 2 cm de largo. pe-
rianto de 2 mm diám., lóbulos ciliados; amentos
femeninos de hasta 3 flores, de 1—1.8 em de largo,
raquis glabrescente; frutos solitarios o en grupos de
2 a 3, sésiles o sobre un pedünculo 4—5 mm de
largo; cúpula hemisférica, de 1.3-1.7 ст diám.,
con las escamas pubescentes, las superiores adpre-
sas, ápice agudo o redondeado, bases engrosadas;
bellota ovoide, pared interna del pericarpo glabra,
de 1.5-1.8 cm de largo, de 1.2-1.3 cm diám.,
cluída en la cúpula un tercio de su largo. Figura 3.
in-
Reconocimiento.
noce porque sus ramillas у el envés de las hojas
Quercus glabrescens se reco-
poseen escasa pubescencia, y porque el borde de
ésta tiene dientes mucronados ubicados en la parte
cal.
Distribución y hábitat.
dos de Distrito Federal, Hidalgo, Estado de México,
Oaxaca, San Luis Potosí, Tlaxcala y Veracruz. En
bosque de Pinus, Pinus-Quercus y en bosque me-
sófilo de montaña, se asocia con Alnus, Abies, Fra-
en altitudes de 2450—
ap
—
=.
En México en los esta-
Clethra y Cornus,
Florece de febrero a junio y fructi-
Encino, encino blan-
No se conocen usos de esta especie; sin embar-
go, De la Paz (1985) la recomienda para la elabo-
ración de mangos de herramienta, pisos industria-
les, cercas, tarimas, construcciones pesadas y para
construcciones en general donde se requiera resis-
lencia.
Ejemplares examinados. MÉXICO. Estado de Méxi-
co: oo
414 (EN CB): юн de ‘la Joya A Ale lealica. |
s.n. (CHAPA, ENCB, MEXU); Ozumba de Alzate, Techi-
ies o, W del Popoc atepetl, Boyas 576 (MEXU); ' ехсосо,
1 al SE de San Pablo Ixayoc, Rzedowski 24171 (Е
Pablo Ixayoc, Е:
X l); Cañada al SE de San
Rafael, Barrios 19 (CHAI ^. ENC B. AEXU); alrededores
s, Rzedowski 31890 (CHAPA, ENCB,
1 km al E de San Rafael E. Rzedowski
32452 eid ); base del Iztaccfhuatl, 4 km al SW de San
Rafael, Vera 114 (ENCB); Villa del Carbón. San Jerónimo,
Martínez 39015 (MEXU
—
4. Quercus glaucoides M. Martens & Galeotti,
Bull. Acad. Roy. Sci. Bruxelles, Vol. 10, pt. 1:
209. 1843. TIPO: México. Oaxaca: Galeotti
103 (BR).
Arbol de 4—10 т de alto, con el tronco de 25—40
cm diám., corteza gris; ramillas de 1-3 mm diám.,
de color castafio-rojizo, glabras o pilosas cerca de la
base de las yemas, con lenticelas de 0.5-1] mm de
largo, protuberantes, pálidas; yemas de 1-2.5 mm
de largo, ovoides, obtusas; escamas ciliadas de color
castafio, con los bordes papiráceas, ápices obtusos;
estípulas de 4—5 mm de largo, lineares, pilosas en
los márgenes; hojas jóvenes muy delgadas, no glau-
cas, haz y envés con abundante pubescencia for-
mada por tricomas estrellados largos, antes de la ma-
durez las hojas son glabras; hojas maduras glabras
o con algunos tricomas cerca de la base, glaucas,
lámina (8—)10—13(—15) X (3.5—)4—7.5 em, coriáceas,
oblanceoladas, elípticas u obovadas, ápice obtuso,
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
Figura 3.
redondeado o retuso, a veces subagudo, base cor-
dada, redondeada, a veces ligeramente oblicua, bor-
de engrosado, cartilaginoso, plano o ligeramente re-
voluto, entero, con ondulaciones o con 4 a 7 dientes
anchos y obtusos de cada lado; nervaduras de (4)7
a 12 de cada lado, ascendentes, arqueadas o casi
rectas, se ramifican cerca del borde de la hoja; haz
glabro, de color verde grisáceo, algo lustroso, ner-
vaduras rojizas o amarillentas, la central y primarias
ligeramente elevadas; envés glabro o con algunos tri-
comas cerca de la base, epidermis glauca-cerosa,
papilosa, blanquecina, nervaduras elevadas; pecíolos
Quercus glabrescens.—A. Rama. —B. Fruto.
—C. Tricoma. —D. Hoja. (A-C: Robledo 3; D: Chen 2.)
de 2-8 mm de largo, de 1-2 mm diám., de color
castaño. rojizo o negro; flores desconocidas; fruto
anual, en pares o en grupos de tres, sésiles, o en
pedúnculos de hasta 6 cm de largo; cúpula hemis-
férica, de 9-16 mm diám., escamas adpresas, en-
grosadas en la base, con abundante pubescencia,
ápices triangulares; bellota ovoide, pared interna del
pericarpo glabra, de 8-12 mm de largo, de 8-9 mm
diám., incluída un tercio de su largo en la cúpula.
Figura 4.
Reconocimiento. Quercus glaucoides se reco-
Annals of the
Missouri Botanical Garden
Figura 4.
noce por sus hojas de color verde-azuloso con el
envés glabro y glauco-acerado. Кагаз veces, mues-
tra similitud con Q. obtusata, debido a que esta
tiltima puede presentar el envés de las hojas glauco
y casi glabro, y los pecíolos cortos de color obscuro,
características muy frecuentes de Q. glaucoides; se
distinguen porque ésta presenta frutos ovoides,
mientras que О. obtusata presenta por lo general,
frutos globosos.
Distribución y hábitat. En México en los esta-
dos de Jalisco, Michoacán, Estado de México, Gua-
najuato, Querétaro, Hidalgo, Morelos, Puebla. Gue-
rrero, Oaxaca. En bosque de Pinus-Quercus, se le
encuentra asociado con Pinus pringlei, Quercus
conspersa, Q. castanea y Q. obtusata, en altitudes
de 750-1800 m.
Fenología. | Fructifica de junio a agosto.
Quercus glaucoides.—A. Rama. —B. Fruto.
(Romero 4031.)
Nombres populares y usos. Tocuz, encino blan-
co, encino negro, encino roble, encino y roble.
=
Se le usa como leña, para horcones, manufactura
de algunos implementos agrícolas y posiblemente
por el área de distribución se explote como material
celulósico (González, 1986); también se emplea
para elaborar carbón, cabos para herramientas,
puertas de golpe y postes para cercas (Bello & La-
yat, 1987). Las hojas y frutos se usan como forraje:
las primeras se emplean como medicinales (Váz-
quez, 1992).
Ejemplares examinados. MÉXICO. Estado de Méxi-
co: Ocuilan de Arteaga, 1 km al S de Chalma, Muller
9220 (MEXU): Santo Tomás de los Plátanos, La Junta,
Martínez 29358 (CODAGEM, MEXU); Sultepec, km 18
carretera Sultepec-San Miguel Totolmoloya, Torres 384
(IZTA); km 12 carretera Sultepec-San Miguel Totolmolo-
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
ya, Torres 388 (17ГА); Tejupilco, Cuadrilla de López, Salto
del / gua, — 6 е esa de Nanchititla, Mar-
tinez Sont (MEXU); Nanc кещ Rodríguez 201 (INIF):
Tena de González, Luvianos, Breedlove s.n.
(MEXU); Cerro de la Culebra, Luvianos Progreso, Matuda
31472 (CODAGEM, MEXU); Tlatlaya,
Llano Tlatlaya, Martínez 30151 (MEXU);
Valle de d Boege s.n. (МЕХ); V
tínez 2609 (INIF); Valle de Bravo, | . (MEXU):
Tiloxtoc, San Bartolo NW de Valle de [A “Muller 9058
(MEXU); Zumpahuacan, cerca de San Gaspar, Tejero &
Castilla 2862 (IZTA).
Hincón Grande у
Valle de Bravo,
Valle ul apt Mar-
Martínez
Quercus laeta Liebmann, Overs. Kongel.
Danske Vidensk. Selsk. Forh. Medlemmers
Arbeider: 179. 1854. TIPO: México. Casa
Grande, Hartweg 419 (B).
Quercus — Trel., Mem. Natl. Acad. Sei. 20: 61, pl
60. . TIPO: México. Distrito Federal: Contreras,
— poe 5c (B)
Árbol de 6-10 m de alto, con el tronco de 25-
40 em diám.; ramillas de 1-2(-3) mm diám., roji-
zas, glabrescentes, con indumento formado por tri-
comas estrellados y simples largos, con el tiempo
se ennegrece, lenticelas menores de 0.5-1 mm, pá-
lidas, en general no muy abundantes; yemas de 2—
4(—5) mm de largo, ovoides, de color castaño, es-
camas con el dorso pubescente y los bordes cilia-
dos: estípulas 4—6(—7) mm de largo, lineares, pu-
bescentes, persistentes en las yemas apicales; hojas
jóvenes de color verde-grisáceo, haz con tricomas
estrellados dispersos, más abundantes en la ner-
vadura central, envés muy pubescente, pubescen-
cia formada por tricomas estrellados con sus ramas
extendidas; hojas maduras de color verde obscuro,
coriáceas, decíduas, elípticas, oblonga-lanceoladas,
oblanceoladas u obovadas, lámina 5.5-12(-14)
(2-)3-5.5(-8.5) cm, ápice obtuso o agudo, base re-
dondeada, cordada, a veces cuneada, borde revo-
luto, engrosado, ondulado o dentado, a veces en-
tero, con 5 a 7 dientes en cada lado, en las dos
terceras partes superiores de las hojas, dientes ob-
tusos, con mucrones engrosados y curvos; nerva-
duras de 7-11(—12) en cada lado, ascendentes, cur-
vas о casi rectas, se prolongan hasta los dientes. y
se dividen en la mitad o cerca del borde de la hoja:
haz lustroso, glabro o con tricomas estrellados uni-
formemente distribuidos y tricomas simples en ner-
vadura central, ésta principalmente pubescente en
su base, nervaduras impresas, envés pubescente
con tricomas estrellados largos, por lo general con
sus ramas extendidas, a veces se enredan un poco
en su base, estípite corto, tricomas glandulares es-
casos; epidermis ligeramente ampulosa, papilosa,
pecíolos de (2-)5-10(-13)
mm de largo, de 1-2.5 mm diám., pubescentes, en-
nervaduras elevadas;
sanchándose en la base; flores desconocidas: fruto
аши, solitario о en grupos de 3, еп pedtinculos de
.5(—4—7) em de largo; cúpula hemisférica de 8—
11 cm diám., escamas pubescentes, principalmente
en la base, ésta engrosada, ápices triangulares, ob-
tusos; bellota ovoide, de 6-13 mm de largo, de 6—
7 mm diám., con una tercera parte o la mitad de
su largo incluída en la cúpula. Figura 5.
Reconocimiento. Quercus laeta se reconoce por
sus hojas con el borde mucronado v envés con pu-
bescencia formada por tricomas estrellados con las
Esta
muestra similitud con Q. obtusata v se distingue
ramas extendidas y estípite corto. especie
porque posee en el envés de las hojas abundantes
tricomas glandulares y escasos tricomas estrellados
con las ramas enredadas entre sí; además el tamafio
de sus hojas es mayor.
Distribución y hábitat. En México en los esta-
dos de Aguascalientes, Coahuila, Distrito Federal.
Durango, Guanajuato, Hidalgo, Jalisco. Estado de
México, Michoacán, Nayarit, Nuevo León, San Luis
Potosí, Sinaloa y Zacatecas. En bosque de Quercus,
bosque de Pinus—Quercus, en matorral xerófilo y
encinar secundario, se asocia con Q. obtusata, Q.
castanea, Pinus, Abies, Cupressus, Arbutus,
Fraxinus, en altitudes de 2350-2750 m.
Fenología.
Alnus y
Florece en febrero y fructifica de
julio a diciembre.
Vombres populares y usos. Encino colorado.
Su madera resulta difícil de trabajar, aunque se
usa para la elaboración de algunas herramientas:
mas no se considera buena para la elaboración de
muebles. También se utiliza como leña, carbón pos-
tería, horcones, cercas y como material celulósico
(González. 1986).
Ejemplares — ч ICO. Estado de Méxi-
co: Atizapál 3. Zar LIA
Rzedow ski 3258. (ZT А); Тар tepec, km 67 carretera ral
ne ron dtp . Cisneros 5 (IZTA); Denhxi, Martínez
64 ( Tenango del Valle, Rancho Yeca, s.n
т. Ton Sierra de
1574 (VZTA
Nicolás Rome ro-T lazala de F
Quintero 36 ПУТА): km 28 carretera a Tlalnepantla
del Carbón, Rojas & = 5165 (IZ TA): Barrio IV, Rojas
& Romero 3192 (Z
Quercus magnoliifolia Née, Anales Ci. Nat
9: 208. TIPO:
México. Guerrero: entre Chilpancingo y Tixtla,
Née s.n. (MA).
1801 (como magnoliaefolia).
Quercus macrophylla Née, Anales Ci. Nat E 274. 1801.
TIPO: Entre € * ша y Tixtla, Née s. a ace
ат МЕ t E. E Warb., Bull. Mise. : 85.
IPO: México. К эе, de México: — 6360
562
Annals of the
Missouri Botanical Garden
Figura 5.
Árbol de 3-15 m de alto, tronco de 20-50 ст
diám.; ramillas de 2—4 mm diám., glabrescentes,
con tricomas estrellados y glandulares cortos, éstos
de color ámbar-rojizo; lenticelas muy notorias, pá-
lidas, menores de 0.5-1.5 mm de largo; yemas de
1.5-5 mm de largo, ovoides, de color castafio, grue-
sas, pubescentes, principalmente en los márgenes:
estípulas filiformes, de 2-10 mm de largo, pubes-
centes; hojas jóvenes rojizas por la abundancia de
tricomas glandulares, con tricomas estrellados muy
cortos, dispersos; envés más pálido; hojas maduras,
por lo general obovadas o elípticas, lámina (7—)10—
27(-32) X 6-12.5(-21) em, ápice obtuso, redon-
deado o agudo, base atenuada, cuneada, redondea-
da o auriculada, borde engrosado, ligeramente re-
dentado o sinuado-dentado,
voluto, sinuado,
dientes 8 a 14 en cada lado, con un mucrón corto
Quercus laeta.—A. Rama. —B. Fruto. —C. Tricomas. (Urbina 4.)
y doblado en el ápice de cada diente u ondulación;
nervaduras primarias 11 a 20 en cada lado, ascen-
dentes, rectas o ligeramente curvadas, pasan direc-
tamente a formar los mucrones, a veces nervaduras
secundarias los forman; haz verde lustroso, glabro,
con tricomas estrellados y glandulares en la base
de la nervadura central, nervaduras impresas, las
secundarias y más pequefias poco notorias; envés
verde-amarillento, más claro que el haz, con pu-
bescencia abundante о escasa, formada de tricomas
glandulares y estrellados sésiles muy cortos, adpre-
sos a la epidermis, ésta ampulosa y papilosa, ner-
vaduras elevadas; pecíolos de 5-10 Х 1—4 mm,
más anchos de la base, pubescentes, principalmen-
te por el envés de las hojas; amentos masculinos
con numerosas flores, de 6—10 cm de largo, perian-
to membranoso, largamente ciliado, de 3—4 mm
Volume 89, Number 4 Romero et al. 563
2002 Quercus en el Estado de México
у
Figura 6. Quercus magnoliifolia.—A. Rama. —B. Fruto. —C. Tricoma. (Soto 283.)
diám., anteras de 10 a 12, de 1-1.5 mm de largo; gadas cuando maduras, de una a dos terceras partes
flores femeninas solitarias, en grupos de 2 o З ó de su largo incluída en la cúpula. Figura
más, dispersas sobre un pedúnculo de 5-10 em de
largo, fruto anual, la cúpula de 14-25 cm diám., Reconocimiento. Quercus magnoliifolia se re-
escamas obtusas pubescentes, engrosadas en su conoce por sus hojas generalmente grandes, obo-
dorso, ápice más delgado, angostamente triangular: vadas o casi elípticas, con mucrones, haz verde lus-
bellota ovoide, pared interna del pericarpo glabra, troso y nervaduras casi rectas.
de 8-22 mm de largo, de 7-15 mm diám., arru- Distribución y hábitat. En México en los esta-
564
Annals of the
Missouri Botanical Garden
dos de Colima, Guerrero, Jalisco,
Nayarit,
bosque de Pinus, Quercus y Pinus—Quercus, se aso-
Estado de Méxi-
co, Michoacan, Oaxaca y Sinaloa. En
cia con Pinus oocarpa, Quercus hintonii y Q. ellip-
lica; se le encuentra también en encinares pertur-
bados, en altitudes de 1700—2600 m.
Fenología. Florece en febrero y fructifica en
junio y julio.
Nombres populares y usos. Encino, encino ro-
ble, roble, encino amarillo, encino napis, encino
prieto, encino blanco, encino bermejo, encino ave-
llano.
Su madera se utiliza como leña, carbón, postería,
horcones, para fabricar mangos para herramientas,
bancos, vigas, postes para corral y para extraer ce-
lulosa. Su follaje sirve para fabricar techos rústicos
de viviendas del campo y sus frutos como forraje
para cerdos (González, 1986; Bello € Labat, 1987;
'ázquez, 1992).
MÉXICO. Estado de Méxi-
co: Amatepec, Amatepe cercanías, Matuda 29822
(CODAGEM); Clachic hhilpan, 7 km al NE de Amatepec,
1049 (INIF); Las Trojas, Rodriguez 185 ——
Ixtapan del Oro, cerro ubicado al SW de la presa Tiloxto
d 850 (INIF); Cerro el € vin nco, al E de la —
Valle de аа Estrada 852 (INIF); Ocuilan де Artea-
ga Ex Arteaga, cerro la — Nah 1179 (INIF); Nuevo
Santo Tomás de los — »8. La Junta, Matuda 29356
(CODAG EM); Sultepec, Las Tinajas, Кейш s.n. (СНА-
РА, 12 carre s ra Sultepec-San Miguel Totol-
moloya, Torres 90, 110, 194 (IZTA); km 18 carretera Sul-
2 (IZTA); Tejupilco, Nanchititla,
JAG EM): Canada de Nanchititla, García
454 (CODAG En ENCB); Nanchititla, García s.n. (INIF);
proximidade % de Tenería, Guízar 155 (CHAPA, ENCB);
camino de Almoloya de la Granadas, Guízar 351 (ENCB);
Cañada de Nanchititla, Soto е Moreno 183, 283 (IZTA);
Popup a ra бе zález, 70 carretera federal 104,
: km 65 de 1
Ejemplares —
—
Castilla & Tejero 1796 (VZT hs km 3-4 carretera Zacual-
pan-Mamatla, Fragoso 339 (17ГА).
7. Quercus obtusata Humboldt € Bonpland, Pl.
Aequinoct. 2: (26). Pl. 76. 1809. TIPO: Méx-
ico. Michoacán: Ario de Rosales, Bonpland
4329 (isotipo, B!).
Bull. Misc. Inform.: 88. 1939.
Quercus atringlans Warb.,
TIPO: México. Estado de México: Temascaltepec de
González, Hinton 6549 (K).
Arbol de 6-20 m de alto, tronco de 40-60 em
diám. o más, con la corteza gris, escamosa; ramillas
—)2—3 mm
indumento formado por tricomas glandulares
glabrescentes, rojizas a gris o negras, de (1
diám.,
y estrellados, lenticelas pálidas, de hasta 2 mm de
argo; yemas ovoides, de (1.5—)2—4(—5) mm de largo,
de color castafio obscuro, escamas pubescentes en
los márgenes; estípulas lineares, de 5-8 mm de lar-
go, membranosas, pubescentes; hojas jóvenes con el
haz rojizo por la abundancia de los tricomas glan-
dulares, envés amarillento, densamente pubescente,
indumento formado por tricomas estrellados entre-
lazados; hojas maduras decíduas, gruesas y coriá-
ceas, rugulosas, obovadas a — obovadas о
elípticas, lámina (4—)6-17(-22) х (2—)3—8(—11) em,
ápice obtuso o anchamente м а veces algo
agudo, borde engrosado, revoluto, dentado, sinuado
o dentado-sinuado, con З a 9 dientes u ondulaciones,
que frecuentemente se distribuyen desde el ápice
hasta la base de la hoja, cada diente u ondulación
termina en un mucrón corto, que se dobla junto con
el borde revoluto, nervaduras de 7 a 12 en cada
lado, ascendentes, rectas o ligeramente curvas, cada
nervadura secundaria pasa a formar un mucrón que
coincide con el ápice del diente u ondulación, a
veces el mucrón se forma entre ondulaciones; haz
verde lustroso, glabrescente, con tricomas simples y
estrellados dispersos, más abundantes en la base de
la nervadura central, nervaduras primarias impresas
o ligeramente protuberantes; envés verde-amarillen-
to, opaco, con pubescencia dispersa formada por tri-
comas estrellados con estípite muy corto, a veces
muy escasos, de aproximadamente 8 rayos, enreda-
dos entre sí o algo extendidos, de pocos a abundan-
tes tricomas glandulares color ámbar o rojizos, a ve-
ces se forman gotas de exudado, epidermis
ligeramente ampulosa y papilosa; pecfolos glabres-
)-11(-15) mm de largo, de 1-2 mm
centes de (3—
diám., a veces de color obscuro; amentos masculinos
de 3 cm de largo, con muchas flores distribuidas a
lo largo del raquis, perianto de 2 mm diám., larga-
mente pubescente, 6 estambres, anteras de 1 mm de
largo, filamento de 1 mm de largo; amentos feme-
ninos de 3 a 6 o más flores distribuidas en la mitad
distal de un pedúnculo de 2-3.5 cm de largo, pu-
bescentes; fruto anual, solitario o en grupos de 2 o
3 ó más, pedünculos de 1.8—3.5 mm de largo; cú-
pulas hemisféricas, de poco a muy profundas, de
12-18 mm diám.; escamas muy pubescentes, ápice
agudo, bases engrosadas; bellota globosa, a veces
cilíndrico-ovoide, con la pared interna del pericarpo
glabra, de 6-20 mm de largo, de 11-19 mm diám.,
hasta un tercio de su largo incluída en la cúpula.
Figura 7.
Reconocimiento. | Quercus obtusata se reconoce
por sus hojas obovadas, con mucrones robustos que
se doblan hacia el envés, éste con abundantes tri-
comas glandulares y escasos tricomas estrellados
con las ramas enredadas entre sí; las bellotas son
globosas. Quercus obtusata muestra similitud con
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
565
Figura 7.
Q. laeta la cual se reconoce por sus hojas con el
borde mucronado y envés con pubescencia formada
por escasos tricomas glandulares y abundantes tri-
comas estrellados con las ramas extendidas y estí-
pite corto.
Distribución y hábitat. En México en los esta-
dos de Guanajuato, Hidalgo, Jalisco, Estado de
México, Michoacán, Morelos, Nayarit, Puebla, San
Luis Potosí y Zacatecas. En bosques de Pinus.
Quercus obtusata.—A. Rama. —B. Frutos. —C. Tricomas. —D. Hoja. (Torres 195.)
Quercus y Pinus-Quercus, se asocia con Pinus
leiophylla, P. montezumae, P. michoacana, Cupres-
sus, Quercus candicans, Q. urbanii y Q. glaucoides;
también se le encuentra en bosque mesófilo de
montaña y pastizal con matorral xerófilo de Acacia
y Opuntia. Es frecuente en encinares perturbados.
en altitudes de 1430-2850 m.
Fenología. | Florece de abril a mayo y fructifica
de agosto a octubre.
566
Annals of the
Missouri Botanical Garden
Nombres populares y usos. Roble, encino, en-
cino prieto, encino calicahuac, encino negro, en-
cino cosahuicahuatl, encino blanco, encino roble,
encino rojo, encino chino, tocuz, uricua, charari.
Se utiliza para leña, carbón, postes para cerca,
implementos agrícolas, horcones, cabos para herra-
mienta, curtir pieles; la corteza tiene usos medici-
nales, y como material de construcción (Bello &
Labat, 1987).
—— duri dus an MÉXICO. Estado de Méx
Acambay, c ondingá, Estrada 1443 (IZTA); Almo-
loya de Andrés, Rodríguez s.n. (INIF):
Coatepec de Harinas, 4 km al E de Coatepec de Harinas,
artínez 2697, 2840 (INIF); Donato Guerra, El Mirador
de Donato Guerra, Toluca-Valle de Bravo, Rojas & Ro-
mero 3242 (IZTA); Ocuilan de Arteaga de Arteaga, Loma
de Fuego, Gutiérrez 403 dd cerro de La Llovizna, Nah
1180, 1181, 2340 (INIF); $
ded INIF); km 12 carretera Suhepee-San Miguel Totol-
— — 195 (IZTA); Tejupilco, N
pm 200 (CHAPA, INIF); Temascalte pec de (
Ype ricones, Hinton 6549 (INIF); Estancia Vieja, 10
S de Temascaltepec de González, Moreno 168 (INIF); Real
г Arriba, Rocha s.n. 0 NIF); Tenancingo, Las Cumbres,
Rodriguez 31 (INIF); Las Cumbres, Rodriguez s.n. (CHA
PA, INIF); Te mange de n popula, Rancho San Luis Acul-
co, Hinton 17989 ; Tepozotlán, Cerro de pie Ca-
be маз, Madrigal & Vela s s.n. (INIF); Texcoco, 1 km al SE
de San Pablo Ixayoc, zedowski 24193 (INIF); Valle de
Bravo, 2 km al E de Pipioltepec, Pineda & Ochoa о
(INIF); Godines Tehuaste — Rodriguez 205, 206 (IN
Villa del Carbón, km 28 carretera Tlalnepantla—Villa de i
— Rojas & Romero 3165 — Las Golondrinas,
a 1000 m del límite con Hidalgo, Rojas & — 3306.
3310 (IZTA); Villa Nicolás Romero, 3 km al W de Ca-
huacán, Rzedowski 32631 (INIF); Zacualpan, Subestación
Coronas, Jiménez 13 (IZTA
—
8. Quercus peduncularis Née, Anales Ci. Nat.
3: 270. 1801. TIPO: México.
Acapulco, Née s.n. (MA).
Entre México v
Querc P — — E. E Warb., AD Misc. — : 87.
. TIPO: México. Estado de México: Temascal-
ài pec doi Conzale ‘z, Hinton — (holotipo, K no vis-
to).
Árbol de 4 m de alto; ramillas de 2-2.5 mm
diám., glabrescentes, en un principio con un denso
tomento formado por tricomas estrellados estipita-
dos con rayos largos; lenticelas visibles sólo en ra-
millas viejas, de color claro, de hasta 1 mm de
largo; yemas de 3—4 mm de largo, ovoides, con las
escamas obtusas, ciliadas en los bordes y dorso;
estípulas de 5-7 mm de largo, subuladas o lineares,
con tomento largo o glabrescente, persistentes aún
en las hojas maduras; hojas jóvenes muy delgadas,
de color verde obscuro, haz con fina pubescencia
formada por tricomas estrellados cortos, algunos
argos, envés densamente pubescente, de color cla-
ro; hojas maduras subcoriáceas, oblanceoladas,
elípticas u obovadas, lámina (4-)7-10 X (2.5—)3—
5 cm, ápice obtuso o agudo, base atenuada, redon-
deada o subcordada, a veces asimétrica, borde car-
tilaginoso, revoluto, casi siempre dentado, con 7 a
10 dientes mucronados a cada lado, desde la base
de la hoja, nervaduras 9 a 14 en cada lado, ascen-
dentes, casi rectas, pasan directamente al diente,
se ramifican cerca del borde; haz lustroso, de color
verde obscuro, glabro, excepto la base que presenta
tricomas estrellados parecidos a los del envés, ner-
vadura central algo convexa, las primeras impresas;
envés con indumento blanquecino formado por tri-
comas estrellados estipitados, los rayos largos y
extendidos que se entrecruzan con otros, epidermis
ampulosa y papilosa, nervaduras pálidas, convexas;
pecíolos de 2
cidas; fruto anual, solitario o en pares sobre pe-
mm diám.;
—5 mm, pubescentes; flores descono-
dúnculos de 3.5—4 em de largo, de
cúpula hemisférica de 12-13 mm diám.,
pubescentes, sus ápices glabrescentes, agudos o
escamas
acuminados; bellota ovoide, pared interna del pe-
ricarpo glabra, de 8-13 mm diám., de uno a dos
tercios incluida en la cúpula. Figura 8.
Reconocimiento. Quercus peduncularis es una
especie muy escasa en la entidad; sin embargo se
encuentran comunidades en donde esta especie es
abundante. Se reconoce porque el envés de sus ho-
jas es tomentoso y blanquecino y los pedúnculos
de los frutos son largos.
Distribución y hábitat. Еп México en los esta-
dos de Colima, Chiapas, Guerrero, Jalisco, Estado
de México, Michoacán, Oaxaca y Veracruz, también
en América Central. Bosque de Pinus—Quercus, se
asocia con Alnus, en altitudes de 2550 m.
Fenología. | Fructifica en julio.
Nombres populares y usos. Encino, roble, en-
cino blanco, encino avellano, encino zopilote negro,
mazcahuite.
La madera se emplea como carbón, lefia y ma-
terial celulósico para papel (González, 1986).
Ejemplares examinados. MÉXICO. Estado de Méxi
co: Te mascaltepec de González, San Lucas, Hinton 6376
Te mascaltepec de González, Hinton 67331
(MEXU): NE de San Francisco Oxtotilpan, Orozco 268
(IZTA).
—
—
Quercus rugosa Née, Anales Ci. Nat. 3: 275.
1801. TIPO: México. Estado de México: Huix-
quilucan y Ocuilan de Arteaga, Née s.n. (MA).
Acad. Sci. 20: 77,
Tlalpu-
Mem. Natl.
: México. Michoacán:
Quercus conglomerata Tre ^
pl. 106. 1924. ТІР
xahua, Hartweg 429 е ).
Quercus decipiens Mart. & Gal., Bull. Acad. Roy. Sci. Bru-
xelles, Vol. 10, pt. 1: 214. 1843. Syn. nov. TIPO
Méx xico. Mirador, Galeotti 131 (BR).
Volume 89, Number 4
Romero et al.
Quercus en el Estado de México
567
Figura 8.
Quercus reticulata Humb. & Bonpl., Pl. Aequinoct. 2:
40(35). pl. 86. 1809. TIPO: México. Guanajuato:
Santa Rosa, Bonpland 4408 (B).
`
Arbol de (5—)10—30 m de alto, con el tronco de
hasta 1 m 6 más diám.; ramillas gruesas, de 1.5
mm diám., glabrescentes, de color gris-castaño, tri-
comas del indumento estrellado-estipitados, que se
ennegrecen con el tiempo; lenticelas hasta de 1
mm, en las ramas viejas más grandes y protube-
rantes; yemas de 2-5 mm de largo, ovoides, de col-
or café-rojizo, escamas coriáceas, pubescentes; es-
típulas lineares u oblanceoladas, decíduas. de 5—7
Ç: К
NES -
"уо;
КАЛД
убу
ЖОНИ
oo К
a)
0.4mm
Quercus peduncularis.—A. Rama. —B. Fruto. С. Tricomas. (Hinton 6331.)
mm de largo, piloso-sedosas; hojas jóvenes con el
haz rojizo y abundantes tricomas glandulares, vena
principal muy pubescente, envés con densa pubes-
cencia pálida, epidermis totalmente cubierta de tri-
comas glandulares amarillentos, nervadura con tri-
comas glandulares rojizos; hojas maduras deciduas,
coriáceas, rígidas, con frecuencia cóncavas, muy
-
rugosas, elípticas, elíptico-obovadas, obovadas «
casi suborbiculares, lámina (3.2-)5-15(-19) X 2—
8-10) em, ápice obtuso, a veces agudo, base re-
dondeada o cordada, borde sinuado o dentado, en-
grosado, plano o revoluto, dientes de 5 a 12 en cada
568
Annals of the
Missouri Botanical Garden
lado distribuidos en las % partes superiores, agudos
u obtusos, con mucrón corto hasta de 1-2 mm de
largo; nervaduras de 7 a 12 en cada lado, rectas o
curvadas, pueden pasar directamente al mucrón о
ramificarse antes y las ramificaciones pasar a los
mucrones adyacentes; haz algo lustroso, verde o
grisáceo, glabro o con pocos tricomas estrellados y
tricomas glandulares, principalmente en la nerva-
dura media, nervaduras central, primarias y secun-
darias impresas, las más finas elevadas formando
un retículo pálido; envés densamente pubescente o
con tricomas esparcidos; pubescencia formada por
tricomas estrellados con un estípite corto, sus rayos
enredados, tricomas glandulares vermiformes de
color ámbar o rojizos; epidermis ligeramente o mar-
cadamente ampulosa, papilosa; pecíolos pubescen-
tes, de (35-10-13) mm de largo, de 1-3 mm
diám., rojizos; amentos masculinos de 15-20 mm
de largo, pubescentes, perianto campanulado, de 2
mm diám., bordes ondulados y largamente ciliados,
anteras glabras; amentos femeninos con 15 a 20
flores en pedtinculos pubescentes; fruto anual, so-
litario o en grupos de 2—5(-8), en pedúnculos de
(0.8-)1.5-5(-8) cm de largo; cúpula hemisférica,
de 8-15 mm diám.,
engrosadas en la base, pubescentes en el dorso, de
de 6-10 mm de alto, escamas
color café rojizo; bellota ovoide o angostamente
elíptica, pared interna del pericarpo glabro, ápice
agudo de 9-28 mm de largo, de 7-14 mm diám.,
incluída una tercera o hasta la mitad de su largo.
Figura 9.
Reconocimiento. Quercus rugosa se reconoce
por sus hojas coriáceas y cóncavas, los mucrones
largos y envés con tricomas glandulares y depósitos
de mucílago. Quercus rugosa puede confundirse
con Q. obtusata; ésta última se distingue por la
presencia de mucrones callosos en los márgenes de
las hojas, mientras que Q. rugosa los presenta pro-
minentes de hasta 2 mm de largo.
Distribución y hábitat. En México en los esta-
dos de Aguascalientes, Coahuila, Chihuahua. Dis-
trito Federal, Durango, Guanajuato, Hidalgo, Jalis-
co, Estado de México, Michoacán, Puebla, Veracruz
y Zacatecas. En bosques de Pinus—Quercus, de Pi-
nus, de Quercus, de Abies, en matorral xerófilo, en
encinares perturbados y cultivos agrícolas; se aso-
cia con Pinus letophylla, P. oocarpa, P. teocote, P.
pseudostrobus, Quercus castanea, Q. candicans, Q.
laurina y Q. altitudes de 1700-
3500 m.
Fenología.
crassipes, en
Florece en agosto y fructifica de
septiembre a noviembre.
Nombres populares y usos. Roble, doga, encino,
encino roble, encino quebracho, encino hojarasco,
encino negro, encino cuero, encino blanco, encino
asta, encino avellano, encino miel, encino blanco
liso, encino quiebra hacha.
La corteza se utiliza en el tratamiento de la di-
sentería, dolor de muelas, hemorragias y fortaleci-
miento de los dientes; usada junto con las hojas
sirve para preparar la infusión que se utiliza para
el tratamiento contra el dolor muscular, la tos (Chi-
no & Jacques, 1986). Se sabe también que las be-
llotas sirven para elaborar café o pueden consu-
mirse tostadas; también se usan como forraje
(Vázquez, 1992). La madera se utiliza para la ob-
tención de papel y como leña (González, 1986).
Ejemplares examinados. MÉXICO. Estado de Méxi-
co: Acambay, cerro Detiña, Estrada 1467 (IZTA); Aculco
de Pyle km 117 carretera a Toluca, Rojas & Romero
2150 (IZTA); Axapusco, Cerro del Tepayo, Jaltepec, Ven-
tura 675 (INIF); Amanalco de Becerra, San Jerónimo
Amanalco, Ventura = (ENCB); alrededores de Amanalco
le Becerra, Villa 163 (INIF); Amecameca, San Antonio
Zoyatzingo, Avila s „ (МЕ); cerro 7 — Román
109 (IZTA); pee cerro Chiluc › km al SW de
Ciudad Adolfo López Mateos, Райла: к (С HAPA, INIF);
Coyotepec, Sierra de Alcaparrosa, Fernández 134 (ПУТА);
parte alta de la Sierra de Alcaparrosa, 6 km al W de
Coyotepec, Rzedowski 37251 (ENCB); Chapa de Mota,
cercanías del — astronómico, Rojas & Romero
2796 (IZTA); Donato Guerra, Mirador de Donato Guerra,
Rojas & ау 3243 (IZTA); Ecatepec de Morelos, Sierra
de Guadalu wi Suárez 11 (VZTA); — —
Matuda 29023, 29026, 29029 (CHAP: - e
dalgo, La Cima, Gómez 1 (IZTA); cercanías p BE
Hidalgo, Matuda 28621, 32613 (CODAGEM); ы
Jilotepec, Matuda 26668 (CODAGEM); cerro de Jilotepec,
Matuda 27821 (CODAGEM); Llano Grande, 6 km a
de Canalejas, Rojas & Romero 3142 (IZTA); Santa Ana
e Las Manzanas, Hernández 13 (IZTA); Jiquipil-
21 carretera Naucalpan—Jiquipilco,
(ZT A): Ocuilan de Arteaga, Ocuilan de зправа: Martínez
31748 (IZTA); cerro de La Llovizna, Nah AM-5 (INIF);
— de Gómez Farías, Rancher i Buenavista, 8 km al
К de San Marcos, Fuentes АШ-32 y АШ-33 (MEXU); San
n de las Pirám dr cerro Gordo, € wes E Tejero
s.n. (ENCB): Sultepec, Real de Abajo, Машаа 29
м APA); Las Trojes, — 180 (INIF); енен
erro Gordo, Castilla & Tejero 604 (IZTA); Pagos de
popula, cerca de Amecameca, Matuda 18839 (CODA-
E
GEM): Teoloyucan, Sierra de Alcaparrosa, Núñez 43
(IZTA); эж к Sierra de Alcaparrosa, Núñez 2176
(IZTA); Texcoco, 28 km al E de Texcoco, sobre brecha
maderera hacia Tlaloc, García s.n. (INIF); 10 km al E de
Texcoco, al principio de la brecha maderera, García s.n.
A); cerro Tetzc ‘ultzingo, 8 km al E de Texcoco, Pu-
lido 428 (ENCB); 13 km al E de Texcoco, carretera a San
Dieguito, Ricafo 28 (CHAI A): Santa María Tecuanulco,
Vengura 877 (CHAPA, Ps В); Timilpan, Sierra de San
André Cam wacho 288, 397 (IZTA); Tlalmanalco, x de
Tialmanaleo Medellín n. (INIF);; 3 km al SW de San
Rafael Tlalmanalco, — 1033 (ENCB); Tlazala de
Fabela, 2 km al SE de Tlazala de Fabela, Román 289
(ENCB); Valle de ee entre Amanalco de Becerra
Valle de Bravo, Matuda 28628, 28630 (MEXU); San EA
San José Allende, Rojas & Ro-
~
T
Alle ande, alre de dores de S
Romero
Volume 89, Number 4 et al.
2002 Quercus en el Estado de México
569
Figura 9. Quercus rugosa. —A. Rama. —B. Fruto. —C. Tricomas. —D. Rama. (A-C: Román 295; D: Camacho
60.)
mero 2057 (ZTA); Villa del Carbón, Predio Piedra Azul. 215. 1801. TIPO: México. Guerrero: cerca de
Ávila s.n. (INTE); Villa del Carbón. Matuda 693 (CHAPA); Tixtla, Née s.n. (MA)
cercanías de Villa del Carbón, Matuda s.n. (CHAPA); Vil- — к id
la Victoria, cercanías de Villa Victoria. Matuda 26620
(CODAGEM).
pem
Árbol de 10—15 m de alto. corteza gris: ramillas
le 2
=
mm diám., rojizas, con la edad cambian a
10. Quercus splendens Née, Anales Ci. Nat. 3: color café claro, con algunos tricomas estrellados
I 8
570
Annals of the
Missouri Botanical Garden
POOR
APP
Re
3
B y EZT TS C
Figura 10. Quercus splendens.—A. Rama. —B.
cerca de los pecíolos y yemas, con numerosas len-
ticelas pálidas, de hasta | mm de largo; yemas de
2 0 3 mm de largo, ovoides, de color castafio rojizo:
estípulas de 5-8 mm de largo, subuladas, con in-
dumento formado por tricomas largos, persistente
en la yema terminal; hojas jóvenes con pubescencia
fina, formada de tricomas estrellados dispersos en
el haz, envés con tricomas simples y estrellados
sobre la nervadura central; hojas maduras decí-
duas, coriáceas, lanceoladas, elípticas, espatuladas,
lámina 7-18 X 2-6 cm, borde ondulado, revoluto,
engrosado, cartilaginoso, ápice obtuso o emargina-
do, base cordada o redondeada; nervaduras de 11
a de cada lado, ascendentes, ligeramente ar-
queadas, dividiéndose cerca del borde: haz glabro
con las nervaduras impresas; envés con pocos tri-
Fruto, —C. Tricomas. (Torres 529.)
comas estrellados pequefios y abundantes tricomas
simples adpresos a lo largo de la nervadura central,
nervaduras prominentes; epidermis glauco-cerosa,
papilosa; pecfolos de 5-9 mm de largo, de 1-2 mm
diám. en su zona más ancha, rojizos o negros; flores
desconocidas; fruto anual, solitario o en grupos de
2 a 3, sésiles; cúpula hemisférica, de 9-16 mm
diám., escamas con los ápices triangulares, obtusas,
canescentes, engrosadas en la base, sobre todo las
basales: bellota oblata, pared interna del pericarpo
glabra, de 7-9 mm de largo, de 11—15 mm diám.,
incluída en su totalidad en la cúpula. Figura 10.
Reconocimiento. Se reconoce por sus hojas ver-
de-azulosas, sin aristas y la pubescencia del envés
que tiende a concentrarse en la nervadura central.
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
Distribución y hábitat. En México en los esta-
Estado de Méx-
ico, Nayarit, Michoacán, Morelos, Sinaloa y Oaxa-
ca. En bosques de Pinus-Quercus y bosque tropical
caducifolio, en altitudes de 1500—1900 m.
Fenología. | Fructi
dos de Durango, Guerrero, Jalisco,
Fructifica en julio y agosto.
Nombres populares y usos. Encino.
No se tiene reportado algün uso.
MÉXICO. Estado de Méxi-
co: Ama у, cercanfas, Matuda 29824
(ENCB); ec km 18 carretera Sultepec—Salayatla,
Torres 251a (IZTA); km 12 carretera Sultepec-San Miguel
Totolmoloya, Torres 529 (IZTA); Tejupilco, cerro Nanchi-
titla, Matuda 31527 (CODAGEM, ENCB); Mesa de Nan-
chititla, Matuda 32817 (CODAGEM).
E — —
An
11. Quercus acutifolia Née, Anales Ci. Nat. 3:
267. 1801. TIPO: México. Guerrero: cerca de
Tixtla, Née s.n. (holotipo, MA!).
Arbol de 12 m de alto, tronco de 30 ст diám.:
corteza obscura; ramillas de 1-2.5 mm diám., con
costillas, color castaño rojizo, lustrosas, glabras,
lenticelas de hasta 0.5 mm, de color claro abun-
dantes; yemas ovoides, con el ápice agudo, de 4-6
mm de largo, escamas ciliadas en sus bordes: es-
típulas de 5 mm, membranosas, linear-oblanceola-
das, decíduas; hojas jóvenes con tricomas simples
y estrellados dispersos, más abundantes en la ner-
vadura central: hojas maduras subcoriáceas, rígi-
das, verdes, envés más pálido, angostamente elíp-
ticas, lanceoladas u oblanceoladas, lámina 6.5-
14.5 X 2.3-4.2 cm, ápice acuminado o agudo, base
obtusa o cuneada, frecuentemente asimétrica, bor-
de cartilaginoso, ligeramente revoluto, con 8 a 10
dientes aristados a cada lado, aristas de hasta 5
mm de largo; nervaduras de 8 a 14 de cada lado.
casi rectas, dividiéndose a veces desde la mitad,
las nervaduras primarias y secundarias forman aris-
tas; haz glabro, lustroso, nervaduras impresas. la
central algo elevada principalmente en la base: en-
vés glabro. lustroso, epidermis lisa, no papilosa,
nervaduras elevadas, principalmente la central: pe-
cíolos de 9-22 X
a y obscura; flores
mm, glabros, su base más an-
y frutos desconocidos.
Figu-
ra 11
Reconocimiento. (Quercus acutifolia es muy es-
casa en el Estado de México; se reconoce por sus
hojas dentadas y aristadas y envés casi glabro. Q.
acutifolia muestra similitud morfológica con Q.
conspersa, pero esta última se distingue porque el
borde de sus hojas es entero y por su envés per-
sistentemente glandular y marcadamente amari-
llento por la presencia de las glándulas.
Distribución y hábitat. En México en los esta-
dos de Jalisco, México y Michoacán. Se le encuen-
tra en bosque mesófilo de montaña, en altitudes
menores a los 2200 m
Fenología. Florece en enero; su fructificación
es desconocida.
Nombres populares y usos. Encino sencillo, en-
cino blanco, encino laurelillo, encino tepezcohuite.
La madera se utiliza como leña o para elaborar
carbón, postería y pulpa para papel; se recomienda
para chapa fina, muebles de alta calidad ebanística,
pisos lambrines, cajas para empaque, mangos y ca-
bos para herramienta (De la Paz, 1976).
MEXICO. Estado de
México: Valle de Bravo, en parte húmeda, Boege
1786 (MEXU); Rancho Rincón Grande
Muller 9095 (MEXU).
Ejemplares examinados.
Piloxtoc,
12. Quercus candicans Née, Anales Ci. Nat. 3:
277. 1801. TIPO. México.
cuaro, Pringle 3955 (MA).
Michoacán: Pátz-
Árbol de hasta 15 m de alto; tronco hasta de
ramillas (1.5-)2-3(-3.5) mm diám.,
abundante pubescencia amarilla que disminuye
m diám.: con
con el tiempo, con lenticelas de 1-3 mm de largo,
visibles en ramillas con la pubescencia disminuida;
5(—6) mm
yemas ovoides de color castaño, de (2-)3—
` largo, con escamas pilosas; estípulas lineares de
~
10-15 mm de largo, decíduas; hojas jóvenes algo
lustrosas, haz con abundantes tricomas estrellados
cortos y tricomas simples dispersos, envés con pu-
bescencia densa blanca, semejante a la de las hojas
maduras; hojas maduras coriáceas y gruesas, obov-
adas, lámina (3.5-)5-19(-23.5) X 3-11(-14) cm,
ápice obtuso o agudo, aristado, base subcordada a
truncada o estrecha hacia el pecíolo, borde revo-
luto, cartilaginoso, conspicuamente dentado o dien-
tes mal definidos, hasta con 25 aristas de cada lado
distribuidas en las dos terceras partes superiores
de la hoja, aristas hasta de 5 mm de largo; nerva-
duras 8 a 14 cada lado, rectas o ligeramente as-
cendentes pasando directamente al diente; haz lus-
verde tricomas
estrellados y muy dispersos, pero abundantes en la
troso, de color obscuro, con
nervadura central cerca del pecíolo, las nervaduras
más finas forman un retículo blanco, nervaduras
central y primarias impresas o ligeramente eleva-
das; envés con pubescencia densa blanca que con
el tiempo sé amarillenta, formada con tricomas es-
trellados con muchos rayos, sésiles y pueden pre-
sentar un corto estípite; epidermis ampulosa, pap-
Поѕа. nervaduras elevadas; pecíolo (3—)10—15(—-20)
0.5-1.5
illa a rojiza; amentos masculinos laxos, perianto de
mm, con abundante pubescencia amar-
2.5-3 mm diám., pilosos en la parte externa y en
572
Annals of the
Missouri Botanical Garden
Figura 11.
el lugar de inserción de los estambres, anteras ex-
ertas de 1.5 mm de largo apendiculadas, filamentos
de 2.5 mm de largo; fruto anual o bianual, solitario
o en pares sobre pedünculos de 15 mm de largo,
pubescente; cúpula hemisférica de 19-23 mm
diám., borde recto, escamas gruesas, con pubes-
cencia corta y muy abundante, ápices redondeados
argo, de
a agudos, glabros; bellota de 20 mm de
17 mm diám., anchamente ovoide, incluida un ter-
cio de su largo en la cúpula. Figura 12.
Reconocimiento, Se reconoce por sus hojas con
Quercus acutifolia.—A. Rama. —B. Fruto. С. Tricomas. (Muller 9096.)
dientes aristados, haz verde lustroso y envés con
pubescencia blanca.
Distribución y hábitat.
dos de Chiapas, Chihuahua, Durango, Guanajuato,
Hidalgo Jalisco, Estado de México, Nayarit, Oaxaca
y Sinaloa, también en Guatemala. Se le encuentra
de manera escasa o abundante en bosques de Quer-
cus, Quercus—Pinus y bosque mesófilo de montaña,
con Clethra, en altitudes de 2000-
En México en los esta-
y se asocia
2600 m.
Fenología. Florece en mayo y fructifica en no-
viembre.
Volume 89, Number 4 Romero et al. 573
2002 Quercus en el Estado de México
Figura 12. Quercus candicans.—A. Rama. —B. Fruto. —C. Tricoma. —D. Hoja. (Rodríguez s.n.)
Encino de asta. en- Méxic Acambay, ladera S del cerro Hoyo de Lobo,
Nombres populares y usos.
cino cenizo, encino papatla, encino blanco, ahu-
amextli.
Su corteza se usa para dolor de muelas; su efecto
dura hasta 15 días. De la Paz (1976) propone que
se utilice para muebles у gabinetes de alta calidad
ebanística, chapa fina, pisos para residencias, mar-
cos para puertas y ventanas, cajas de empaque, cof-
res, mangos y cabos de herramientas e implementos
agrícolas.
Ejemplares examinados. MÉXICO. Estado de
Rojas & Romero 2059-A (IZTA); Coatepec de Harinas,
1 km al E de Coatepec de Harinas, Martinez 2845
(MEXU); 3 km al NE de Coatepec de Harinas sobre el
camino a Agua Amarga, Rzedowski 30356 (ENCB):
Huixquilucan, alrededores de Dos Ríos, Román 317
В); 1 km al NW de Santiago Yancuitlalpan, Rze-
dowski 33489 (ENCB, MEXU); Jilote pee, Cerro de Jil-
otepec, Matuda 29054 (CODAGEM, ENCB, MEXU);
Santa Ana Jilotzingo, Sierra del s Alto, 3 km al
/ de Santa Ana Jilotzingo, Rzedowski 22412 (ENCB,
MEXU — de la Sal, km 106 de la carretera a
вео de la . Martinez 2847 (MEXU):
en ladera des da, Matuda 28814 (CODA(
tH
T
Malinalco,
¿EM
Annals of the
Missouri Botanical Garden
MEXU); Ocuilan de Arteaga, km 18 de la terracería
Ocuilan-Cuernavaca, Cortés 25 (IZTA); 1 km al S
Ocuilan de Arteaga, Muller 9218 MENU js PCR
km 16 carretera Sultepec—San Miguel Totolmoloya, Tor-
res 302 (IZTA); Temascaltepec de González, Comunal
de Tequisquiapan, Huerta € Ramos T-14 (ENCB,
MEXU); El Guajero, rumbo a Real de Arriba, Rojas &
Romero 2370 (VZTA); km 57 carretera Toluca—Temas-
caltepec, Rojas & Romero 2426 (IZTA): ү 51 Toluca-
ев & Romero — (IZTA); Tenan-
i үз nanc ingo, Martinez 65 (MEXU); us de
Yeca, Ávila s.n. (E у B, IZTA); Villa del
Gonz zález s.n. (IZTA); San Jerónimo,
(U); Cerro Bufa, Matuda 29753
2 Guerrero, 5 km al N de Villa —
. (ENCB); Villa Nicolás Romero, km ¢
carretera V illa Ni ‘olds Romero—Tlazala, límite ш -
ipal de Villa Nicolás Romero, — 33 (IZTA): Bar-
rio IV, Rojas & Romero 3191 (IZTA); Cahuacán, Кой аў
& Romero 3191 A (IZTA); 3 km al NW de Cahuacán,
Rzedowski 32626 (ENCB).
epec, Aojas
me tho Y
. Quercus castanea Née, Anales Ci. Nat.
276. 1801. TIPO: México. Hidalgo: Née s.n.
(MA).
Quercus axillaris Fourn. ex Trel., Mem. Natl. Acad. Sci.
20: 176, pl. 352. 1924. TIPO: México. San Nicolas,
Valle de México, Bourgeau 1138 (B).
Quercus pulchella Humb. * Bonpl., Pl.
(38)44, pl. 88. 1809. TIPO: México.
Santa Rosa, ie s.n. (P).
Árbol de 5-15 m de alto, tronco de 40—80 cm
ramillas de (0.5—)1—2 mm diám.,
café claro a obscuro, pubescentes, a veces glabras
Aequinoct. 2:
Guanajuato:
diám.; de color
con varias costillas, lenticelas pálidas de 0.5-1 mm
de largo: yemas de (1—)2—3.5(—7) mm de largo, ovo-
ides o agudas, de color café, con escamas ovadas,
ciliadas en los márgenes y dorso superior, cori-
áceas; estípulas de 5-6 mm de largo, lanceoladas,
de color claro, membranosas, con tricomas largos
muy
principalmente en los márgenes, caedizas
pronto; hojas jóvenes muy parecidas a las maduras,
haz verde con pubescencia principalmente en la
base y nervaduras, envés densamente pubescente;
hojas maduras gruesas, elípticas, elfptico-oblongas,
oblanceoladas o lanceoladas, láminas (1.8—)3—9(—
11.6)
istado, base redondeada a cordada, borde plano o
14(-5.5) em, ápice agudo u obtuso, ar-
ligeramente revoluto, entero o dentado, con 1-7
aristas de hasta 2 mm de largo, en cada lado del
tercio o mitad superior; nervaduras primarias de
(3235-12 en cada lado, ascendentes, se ramifican
y reticulan cerca del borde, generalmente se con-
tinúan en una arista; haz verde claro a obscuro,
lustroso, finamente rugoso por las nervaduras, gla-
bro excepto cerca del pecíolo, nervadura central
elevada principalmente en la base, nervaduras pri-
marias y secundarias impresas, las menores forman
un retículo pálido; envés muy pubescente, la abun-
dancia de tricomas disminuye con la edad, pubes-
cencia grisácea, formada de tricomas estrellado-es-
tipitados, sus rayos rugosos y enredados, epidermis
con ámpulas prominentes, nervaduras conspicua-
mente elevadas, pálidas; pecíolos de (2.5—)4—10(—
se ensanchan en la base, pu-
5) X 0.5-1.5 mm,
bescente o glabros; amentos masculinos de 6 cm
de largo, pedtinculos pubescentes, perianto de 1.5—
2 mm diám., bordes ciliados, anteras de 1.5 mm de
amentos fe-
argo, filamentos de | mm de largo:
meninos con | o 2 flores, pedúnculos de 3-5 mm
de largo, con pubescencia muy corta formada de
tricomas estrellados; fruto anual, 1 о 2 sésiles o
sobre un pedúnculo de 1-7 mm de largo; cúpula
hemisférica de 9-14 mm diám., con escamas algo
engrosadas en la base, ápice obtuso y papiráceo,
pubescentes a casi glabras, de color café rojizo;
bellota anchamente ovoide, pared interna del per-
icarpo lanosa, de 5-15 mm de largo y de 8-11 mm
diám., incluida en la cúpula de un tercio a un me-
dio de su largo. Figura 13.
Reconocimiento. Esta especie se reconoce por
sus hojas aristadas y envés con las nervaduras con-
spicuamente elevadas y reticuladas.
Distribución y hábitat. En México en los esta-
dos de Colima, Distrito Federal, Durango, Guana-
juato, Guerrero, Hidalgo, Jalisco, Estado de Méxi-
Morelos, Nayarit, Oaxaca, Puebla,
San Luis Potosí, Sonora, Sinaloa y Veracruz, tam-
co, Michoacán,
bién en Guatemala. Se encuentra en bosques de
Pinus, Quercus y Pinus—Quercus, pastizal con ma-
torral xerófilo y bosque mesófilo de montaña; es
frecuente encontrarla en encinares perturbados. Se
le ha visto asociada con Quercus obtusata, Q. glau-
coides, Q. conspersa, Clethra, Arbutus y Pinus mi-
choacana, en altitudes de 1900-3500 m.
Fenología. Florece en junio-julio y fructifica
de agosto a diciembre.
Nombres populares y usos. Encino, encino blan-
co, encino negro, encino amarillo, encino rojo, palo
colorado, encino pipitillo, roble, encino prieto, te-
pozcohuite chino, encino chaparro, encino colorado
y aguacatillo.
De la Paz (1982) recomienda su madera para
pisos de residencias, vehículos (de motor y no mo-
torizados), tarimas para carga y descarga, mangos
y cabos de herramienta, implementos agrícolas, y
diversos tipos de recipientes y armazones de con-
strucción.
— ке MEXICO. Estado de Méxi-
co: Acamba ITO — a 1442 (IZTA); AI-
moloya del ш. Can de Andrés, Rodriguez 147
(ENC — | km A NW W cae San Antonio Zoy-
atzingo, Reyes 211 (ENCB, MEXU); Atizapán, vertiente
SE del cerro Chiluca, 6 km al SW de C Pe Adolfo López
Volume 89, Number 4
2002
Romero et
al. 575
Quercus en el Estado de México
Figura 13.
dríguez s.n.)
Mateos, Patiño 335 (ENCB); — nte SE del cerro a hil-
uca, Polaco 214, 335 (ENCB); Cerro de Tigre, al NW de
— Adolfo López Mateos, Ps dowski 32036 (E NE В,
XU); parte alta del cerro Chiluca, cerca de Ciudad
нея те Mateos, Rzedowski 32586 (ENCB, MEXU):
El Oro de Hidalgo, entre el Oro y Villa Victoria, Matuda
28644 (CODAGEM, MEXU); cercanías del Oro, Машаа
28646 (ENCB): ds pa an, entre San Bartolo y Toluca.
е 495 (ЕМ km al N de San Bartolito, García
189 (ENCB); Magdale na, Rzedowski EUN (E NC B
=
Jilotzingo, Sierra del Monte Alto, :
Ana Jilotzingo. Rzedowski 22414 (ENCB): Malinalco. Mal-
Quercus castanea.—A. Rama. —B. Fruto. —C
. Tricoma. —D. Hojas. —E. Inflorescencia. (A-E: Ro-
inalco, Matuda 28776 (ENCB); Naucalpan de Juárez,
Juárez, COTECOCA 15072 (ENCB): al-
San Francisco C himalpa, Rzedowski 20117
(ENCB. ME XU); 8 km al W de San Bartolo, Rzedowski
32243 (ENCB, МЕХ); Oc ile de Arteaga, El Ahuehue-
te, Rzedowski 26922 (ENCB); Otzoloapan, Cerro de Pinal,
Matuda s.n. (CODAGEM, MEXU); Polotitlán, km 147 car-
retera a Toluca, Rojas & Romero 3099 (1; ); San Felipe
del Progreso, entre el Oro y Villa Victoria, Matuda 2864
(ENCB) Nuevo Santo Tomás de los Pl:
Nuevo Santo Tomás de los Plátan
(ENCB): Sultepec, Real de Abajo, а 29300 (ENCB.
МЕХ); 2 km al NE de Sultepec, Rodríguez 166 (ENCB):
km 18 carretera Sultepec-San Miguel Totolmoloya, Torres
576 Annals of the
Missouri Botanical Garden
317 (IZTA); Tejupilco, Nanchititla, Rodríguez 199 (ENCB, — |ustroso, de color olivo, glabro, excepto en la ner-
MEXU); Temascaltepec de González, Tenería, Guizar 499
(MEXU); San Lucas del Pulque, Hinton 6738 (MEXU);
Real de Arriba, Huerta RA-1 (ENCB); Temascaltepec de
González, Mat d — (ENCB); La Labor, Mat uda
2763 (ENCB, MEXU); Estancia Vieja, 10 km al S de Te-
mascaltepec de ( Ludis Moreno 147 (Е NC B); km 4 Val-
le de Bravo-Temascaltepec, dirección a
Pulque, Orozco & Rojas 520 E (IZTA); Barrio la Mesa,
— Ке ны вни Orozco & Rojas 610 A (IZTA); km
alle de B — mascaltepec, dirección a la Albarrada,
* o 284 E 5 km al SW de Temascaltepec de
González, EARS 208. 30 (ENCB); d —
cingo, Martínez 698 (ENCB); Tenancingo, r 697
(MEXU); 10 km : N de Tenancingo, Muller 9194 (Е NC B);
Comunal Real de Arriba, Ramos & Rocha s.n. (MEXU);
Tenango, entre Te e y€ iai m, Martínez s s.n. Pp
А арча 2 km al N de Magú, Cruz 493, 496 (ENCB,
XU); Sierra de Alca aparrosa, Núñez 1710 ZA Cer-
canfas de la Presa Cone “pe ión, Rzedowski 22908 (ENCB);
i Alc — rrosa, 2 km al NNW de
8 (ENCB); Tlatlaya, Tlatlaya
y cercanfas, Matuda 29829 (CODAGEM, ENCB, ME XU}
Valle de Bravo, Valle de Bravo, Martínez 2612 (ENCB);
— buea pec, Rodrigues 204 (ENCB); Villa del
5 'rogreso. Industrial, d is
=!
T
Es
=
€
arbón, 5 al W de I
porta — (E NCB, MEXU); Villa Guerrero, Villa
rero, COTECOCA 15077 (ENCB); Villa Nicolás —
З km al E de Cahuacán, Espinosa 621 (ENCB); Cahuacán,
Germán & Trejo 43 (MEXU); km 32 carretera —
tla—Villa del Carbón, Rojas & Romero 2019 (IZTA); 1 km
2 S de — 'án, Rzedowski 16815 (ENCB); * ualpan,
km 3-4 carretera Zacualpan-Mamatla, Fragoso 340
(IZTA).
14. Quercus conspersa Bentham, Pl. Hartw. 91.
1842. TIPO: Guatemala. Las Casillas, Hartweg
617 (isotipo, NY!).
Árbol de 9 m de alto, tronco de 50 cm diám.,
ramillas de 2-3 mm diám., gla-
bras, de color castafio rojizo; lenticelas numerosas,
de hasta 0.5 mm de largo;
corteza obscura;
blancas, protuberantes,
yemas de 2—3 mm de largo, ovoides, ápice agudo,
con escamas ciliadas en sus bordes, de color cas-
tafio-rojizo; estípulas de 7-8 mm de largo, lineares,
pilosas; persistentes en las hojas jóvenes; hojas jóv-
enes rojizas, haz con tricomas simples y estrellados,
escasos y dispersos, envés con abundantes tricomas
glandulares de color ámbar y estrellados con un
largo estípite y rayos enredados entre sí, abundan-
tes en las axilas de las nervaduras primarias; hojas
maduras gruesas, coriáceas, decíduas, elíptico-ova-
das, lanceoladas u oblanceoladas, lámina (8—)11—
17(-22) X (2-)3—6.5 ст, ápice acuminado o agudo
aristado, base obtusa, subcordada o atenuada, a ve-
ces asimétrica, borde grueso, cartilaginoso, revo-
luto, entero; nervaduras primarias de 7 a 10 en
cada lado, ascendentes, en hojas elíptico-ovadas
son casi paralelas, ramificándose, a veces, desde la
haz
base de la nervadura hasta cerca del borde;
vadura central en la que se observan abundantes
tricomas glandulares obscuros y tricomas estrella-
dos pequefios en la base de la hoja, nervaduras
impresas, envés también lustroso, amarillento por
la presencia de tricomas glandulares color ámbar,
con mechones de tricomas estrellados con estípite
largo y rayos enredados entre sí, concentrados en
las axilas de las nervaduras primarias, epidermis
papilosa, nervaduras prominentes; pecíolos de 8—
15 mm de largo, engrosados en su base, de color
rojizo a negro, en un principio pubescentes, luego
glabros-rugosos; flores desconocidas; fruto bianual,
solitario, sobre un pedúnculo grueso, de 4-6 mm
de largo; cúpula poco profunda, de 14 mm diám.,
borde enrollado; escamas con pubescencia blanca,
adpresas, no engrosadas en la base, ápices delto-
ides u obtusos; bellota comprimida, pared interna
del pericarpo lanosa, de 5 mm de largo, de 10 mm
diám., la mitad de su longitud incluida en la cu-
pula. Figura 14.
Reconocimiento. Quercus conspersa se reconoce
por sus hojas de borde entero, haz lustroso de color
verde-olivo y envés amarillento por la presencia de
tricomas glandulares. Esta especie se confunde con
Q. acutifolia (ver Q. acutifolia).
Distribución y hábitat. En México en los estados
de Chiapas, Guerrero, Jalisco, Estado de México,
Michoacán, Oaxaca y Veracruz, también en Amér-
ica Central. Se le encuentra en bosque de Quercus—
Pinus, y se le encuentra asociado a Pinus oocarpa,
en altitudes menores de 2000 m.
Fenología. | Fructifica de diciembre a febrero.
Nombres populares y usos. Encino, roble amar-
illo, encino rojo, encino colorado, encino pipitillo,
tepozcohuite, encino cáscara, encino blanco.
Su madera es utilizada para la elaboración de
chapa y pulpa para papel (González, 1986).
Ejemplares examinados. MÉXICO. Estado de Méxi-
co: Temasca ерес de González, Pungarancho, Hinton
7570 (MEXU); 0.5 km al SE de Albarrada, 7 km al W de
Temascaltepec de González, Muller 9104 (ME XU); km 15
desviación San Pedro Tenayac, Orozco 144 (IZTA)
15. Quercus crassifolia Humboldt & сас
РІ. Aequinoct., 2: 49, pl. 91. 1809. ТІР
México. Guerrero: Chilpancingo, a
п. (В,
Arbolillo de 2-3 m de alto o árbol de hasta 15
m de alto, tronco hasta de 1 m diám., ramillas de
2-5 mm diám., con abundantes tricomas estrella-
dos estipitados cubriendo totalmente la superficie,
pubescencia amarilla a café que se obscurece y cae
con el tiempo; lenticelas numerosas de hasta 3 mm,
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
577
Figura 14.
claras, visibles sólo en las ramas que han perdido
la pubescencia; yemas de 3-7 mm, ovoides, pilo-
sas; estípulas de 8-12 mm de largo, oblanceoladas,
escariosas, sedoso pubescentes, con el tiempo gla-
bras, persistentes en las yemas, a veces también en
las hojas maduras; hojas jóvenes lanoso-tomento-
Quercus conspersa.—A. Rama. —HB-D. Frutos. (Hinton 7570.)
sas, haz rojizo y envés café, con aristas de hasta 3
mm de largo, indumento formado por tricomas es-
trellados estipitados y glandulares rojizos sobre la
epidermis obscura; hojas maduras por lo general
obovadas, oblongo-obovadas, elípticas, coriáceas,
5-)7-17(-20) х 3.5-10(-11.5) cm, ápice
ámina (4
578
Annals of the
Missouri Botanical Garden
agudo, acuminado u obtuso, aristado, base cordada
o redondeada, borde ondulado o dentado, revoluto,
cartilaginoso, de 3 a 8 dientes en cada lado, ubi-
cados casi desde la base de la hoja. 1 a 10 aristas.
dientes y aristas con frecuencia asimétricos; ner-
vaduras 6 a 12 a cada lado, ascendentes, ramifi-
cándose cerca del borde, pasan directamente a los
dientes y forman las aristas, con frecuencia hacen
—
lo mismo las ramificaciones; haz lustroso, glabro,
excepto la nervadura central que posee abundantes
tricomas estrellados, sobre todo cerca del pecíolo,
nervaduras impresas; envés amarillo anaranjado o
café, lanoso-tomentoso, indumento formado por tri-
comas estrellados con un largo estípite, sus ramas
muy largas y entrelazadas, cubriendo la epidermis
ampulosa y papilosa, nervaduras elevadas; pecíolo
con pubescencia densa formada por tricomas es-
trellados más cortos que los del envés, ensanchado
en su base, de 6-15(-17) X (1-)2-3 mm: amentos
masculinos de 8-9 mm de largo, raquis pubescente,
—
perianto amplio, de 2-2.5 mm diám., anteras cur-
vadas de 2 mm de largo, amarillentas-cafés, apen-
diculadas; fruto anual o bianual, solitario o en
pares, sésiles o en pedúnculos de 3-5 mm de largo:
cúpula hemisférica, de 11-13 mm diám., con las
escamas delgadas adpresas, pubescentes, princi-
palmente en sus ápices, éstos redondeados; bellota
ovoide, pared interna del pericarpo lanosa, de 10—
17 mm de largo, de 8—11 mm diám., incluida en
la cúpula % de su largo. Figura 15
Reconocimiento. Quercus crassifolia se recono-
ce por sus hojas aristadas con el envés lanoso
amarillo, anaranjado o café y ramillas de 1-5 mm
Чат. Esta especie muestra similitud con Q. hin-
tonii; ésta última se distingue porque el envés de
sus hojas posee epidermis lisa y las ramas de los
tricomas estrellados son más largas. En ocasiones
también puede confundirse con Q. dysophylla, pero
ésta última posee hojas de forma ovada, lanceolada
o elíptica, por lo general de borde entero; mientras
que Q. crassifolia casi siempre las presenta obov-
adas o dentadas. Además los pecíolos son más cor-
tos en Q. dysoph ylla.
Distribución y hábitat. En México en los esta-
dos de Chiapas, Guerrero, Hidalgo, Jalisco, Estado
de México, Michoacán, Oaxaca, Puebla, Querétaro,
San Luis Potosí, Tlaxcala y Veracruz, también en
Guatemala. Se le encuentra en bosques de Quercus,
у Quercus—Pinus, se asocia con Pinus leiophylla, О.
laurina y Q. crassipes, y se le ha encontrado en
zonas de regeneración después de incendios, en al-
titudes de 2500-2800 m.
Fenología. | Florece en abril y fructifica en oc-
tubre.
Nombres populares y usos. Encino chicharrón,
encino colorado, encino prieto, encino roble, encino
hojarasco, encino huaje, chi-ka-chi, hoja ancha.
La madera de esta especie se utiliza para la man-
ufactura de implementos agrícolas, para horcones
de casas, mangos de herramientas y leña: el carbón
obtenido de esta especie resulta durable y de buen
peso (Camacho, 1985); los retoños son comidos una
vez cocidos, molidos y mezclados con maíz; junto
con carrizo de monte, sauco, toronjil y zarza se em-
plea para hacer un agua caliente para mujeres des-
pués del parto; la corteza se usa para aliviar dolores
de encías, para curtir pieles y en la preparación de
bebidas de Agave (Pennington, 1969).
Ejemplares — MEXICO. Estado de Méxi-
co: Acambay, Martinez s.n. (ENCB); Aculco de Espinosa,
l farther da (МЕХ): Axapusco, Cerro del Tipayo,
ЕМСВ); Coatepec de Harinas, sobre el cam-
Amar E — ski 30351 (ENCB); Capul-
huac, Rodríguez s.n. CB); El Oro de Hidalgo, Bassoco 0,
Rojas & 3.
оа
— en (ИТА
— Ne B): Peña Desc ‘ani, Min nez s.n. MT XI 2 Jil-
otepec, Matuda 26673 (ENCB); Cerro de Jilotepec, Ma-
"na 29058 (CODAGEM); Santa Ana Jilotzingo, Sierra de
Monte Alto, al NE de S: Jilotzingo, Rzedowski
22411 (ENCB); Temascaltepec de González, San Francis-
co Oxtotilpan, re gion Matlatzinca, Fragoso R. 253 (IZTA);
AN Sierra de San — Camacho 153, 233, 293,
ZTA); 4 km al V cabecera municipal, Gómez
с vo 4 — Tlazala de Fabela, alrededores de
Tlazala de Fabela, Román 275 (MEXU); 2 km al SE de
Tlazala de F al, 290 (ENCB); Villa del —
Cerro de la Cabra, Bringas s.n. (ENCB); 1 km a de
Villa del C у Cruz 169 (ENCB); Cerro La An ar-
tinez s.n. (ME XU): cercanías de Villa del Carbón, Maiuda
—
28767 (E NCB); San Jerónimo, Matuda 29017 (С )DA-
GEM); San Je rónimo, Mao 29160 (ENCB); km 15 Ji-
quipilco-Villa del € n, Rojas & о 3175 (IZTA):
El Plan, Rodríguez s. s.n. - (zi TA 0: Villa N cola А Котего, 5
SW «
—
=
—
wn
7
-
>
— 1831 0 (ENC В, МЕХ); 2 km al NW
huacán, Rzedowski 33755.
16. Quercus crassipes Humboldt & Bonpland,
PI. Aequinoct. 2: 37. Pl. 83. 1809. TIPO: Méx-
ico. Guanajuato: Santa Rosa, Bonpland s.n.
otipo, B!).
—
18-
Árbol de 4—17 m de alto o más, con el tronco
de 0.40—1 m diám.; corteza de placas alargadas о
de color obscuro; ramillas de (5—)1 2 mm diám.,
con pubescencia densa amarilla, formada por tri-
comas estrellados con estípite muy pequeño: len-
—
ticelas hasta de 1 mm de largo, desde pálidas hasta
del mismo color de las ramas; yemas de 1.54.5
mm de largo. ovoides, de color café-rojizo, escamas
coriáceas, bordes ciliados; estípulas de 7-8 mm de
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
Figura 15.
3081; B-D: Camacho 15.
Quercus а —A. Ката. —B. Fruto.
largo. linear-lanceoladas, membranosas pubescen-
tes en el dorso, decíduas; hojas jóvenes con abun-
dante pubescencia amarilla en haz y envés, prin-
cipalmente en la nervadura central; hojas maduras
coriáceas, angostamente elípticas, lanceoladas и
oblanceoladas, lámina 2-9(-10.8) X (0.6) 1-34)
em, ápice mucronado o con arista de 3 mm de lar-
go. base redondeada o subcordada, borde entero.
—С. Tricoma. =). Hoja. —E (A: Romero
. Inflorescencia.
revoluto, engrosado: nervaduras de 10 a 19 en cada
lado, rectas o algo curvadas, formando ángulos casi
rectos, bifureados cerca del margen: haz algo lus-
troso, color verde o grisáceo, glabro o con pequeños
tricomas estrellados dispersos, muy abundantes en
a base de la hoja. nervadura central elevada, las
primeras impresas, las más finas forman un retículo
pálido sobre un fondo verde; envés con pubescen-
579
580
Annals of the
Missouri Botanical Garden
AJ
S SIN
7272»
(LLL
Figura 16.
20.)
cia densa, grisáceo, tricomas estrellados estipita-
dos, con 5 a 6 rayos extendidos, epidermis ampu-
losa; nervaduras ligeramente elevadas: pecíolos
amarillentos o rojizos, pubescentes o casi glabros,
(1-)2-7(-10) mm de largo, de 0.5-1 mm diám.:
4—5.5 cm de largo; flores
con el perianto escarioso de 4 X 3 mm, café-rojizo,
pubescentes; estambres 5, de 3 mm de largo, an-
teras apiculadas; flores femeninas de
amentos masculinos de
o 2 sobre
pedúnculos de 5 mm de largo o menos, de 2—2.5
mm diám.; fruto bianual. solitario o por pares en
pedünculos de 2-8 mm de largo: cúpula hemisfér-
Quercus crassipes.—A. Rama. —B. Fruto.
2»
+
<<
Д ERES
LLZ
—
1ст
у,
ES
^ XS
RE
A Z KE
(LEED SAN: Мах
—
SER
NN
0.5mm
—
|. Tricomas. —
Ноја. (A-C: Romero 2055; D: Acosta
ica, de 11-17 mm diám., à veces invo-
lutos, las escamas engrosadas en la base, pubes-
centes, a veces glabrescentes; bellota ovoide, pared
interna del pericarpo lanosa, de 12—17(—30) mm de
largo, de 8-15 mm diám., cerca de la tercera parte
de su largo incluida en la cúpula. Figura 16.
márgenes
Reconocimiento. Quercus crassipes se reconoce
por sus hojas con el ápice aristado y las nervaduras
que forman ángulos casi rectos: muestra similitud
con Q. mexicana. Esta se diferencia porque el en-
vés de la hoja posee tricomas estrellados con sus
Volume 89, Number 4
Romero et al.
Quercus en el Estado de México
ramas enredadas entre sf, de manera que a simple
vista se observan como puntuaciones.
Distribución y hábitat. En México en los esta-
dos de Jalisco, Guanajuato, Querétaro, Hidalgo,
Colima, Michoacán, Estado de México, Distrito
Federal. Morelos, Tlaxcala y Puebla. Se le encuen-
tra en bosques de Quercus, Pinus-Quercus у Quer-
cus—Cupressus, bosque mesófilo de montafia, mator-
ral xerófilo, en sitios de transición de pastizal a
bosque mixto; se asocia con Pinus pseudostrobus, P.
leiophylla, P. montezumae y P. hartwegii, Quercus
laurina, Q. crassifolia, Q. obtusata y Q. castanea,
en altitudes de 1900-3500 m
Fenología. | Florece en mayo y fructifica de sep-
tiembre a enero.
Nombres populares y usos.
lorado, encino chilillo, encino oreja de ratón y en-
Encino, encino co-
cino laurel.
De la Paz (1982) recomienda su madera para
pisos de residencias, auditorios, museos, almace-
nes, pistas de baile (en forma de parquet y ado-
quín), para chapa fina, muebles y gabinetes de alta
calidad ebanística, lambrín, decoración de estudios
y corredores, cocinas integrales, baúles, canastos,
macetas, cofres y diversos artículos decorativos,
mangos para herramientas, lomos y mangos de ce-
pillos, brochas y de utensilios de cocina, pasama-
nos, huellas (escalones) y descansos de escaleras,
hormas para zapatos y cajas para pianos.
— ip orco MÉXICO. Estado de Méxi-
ulco de Espinosa, Bosque Nado, 10 km antes de
Ad Ds o de Espinosa, Matuda 28731 (ENCB); Almoloya de
Juárez, km 11 de la carretera Toluca-Valle de Bravo, Ks-
trada 346 (ENCB); Amecameca, San Antonio Zovatzingo,
Ávila s.n. (ENCB); 1 km al NE de Santo Tomás Atzingo,
Cervantes 181 (ENCB); 3 km al SW de San Antonio Zoy-
atzingo, DM 528 (ENCB); Cerro Sacromonte, Rzedows-
ki 26695a (ENCB); Capulhuac, Cerro т | km al W
de San Minne Almaya, Rodriguez s.n. (EN Coacalco,
6 km al S de Coacalco, Sierra de Dono “Resins ski
30777 (ENCB); Chalco, Ladera E del Cerro La Tijera,
km al SE de Santa Ana, Pineda Va ҮРҮ, Chicoloa-
pan, 4 km al E de Coatepec, Román 311 (ENCB); El Oro
de Hidalgo, El Oro de Hidalgo, ee 26420, 28647
(CODAGEM); cercanías del Oro, Matuda 28624 (ENCB);
Morelos, 6 km al N de Santa Clara, Ramírez
ENCB); а Fraccionamiento La Herrad-
ura, аа s.n. (ENCB); Dos Ríos, Matuda 28405,
28406 (ENCB); alrededores de Dos ys I1: 327
(ENCB); Ixtapaluca, Río Frío, García 92 ( ied km 42
carretera México-Puebla, Hernández 64. B); Jilote-
Denxi, Martínez 67 (IZTA); Jilotepec, Mud: 26674
Matuda 29052
29065 (ENCB);
o de Jocotitlán, Matuda 30238 (ENCB);
Otimba de ok Farías, 15 km al NE de Texcoco,
Ochoa 40 (ENCB); Cerro Cuixi al E de Santa Bárbara,
Rzedowski 16889 (ENCB); Otzoloapan, Cerro de Pinal,
Matuda 31754, 31870 (CODAGEM); Ozumba de Alzate,
1 S de Ozumba de Alzate, COTECOCA 15036 (ENCB);
—
~=
No
Js
pec,
(CODAGEM, ENCB); Cerro de qo nd
—
Tenango de Tepopula, San Luis Aculco de Espinosa, Hin-
ton 115, 117 (ENCB); 3 km al W de Tenango de Tepopula,
Pineda 731 (ENCB); San Martín de las Pirámides, Cum-
bre del Cerro Gordo, Rzedowski 18801 (ENC В); Temas-
caltepec de González, Cieneguillas, 14 km al W de Te-
mascaltepec de González, Moreno 185 (CHAPA, ENCB):
Temascalapa, Cerro Gordo, Castilla & Tejero 634 (ENCB,
IZTA); Tepetlaoxtoc, 18 km al E de Texcoco, por la car-
retera a Calpulalpan, De la Cruz 15 (ENCB); 14 km al E
de Texcoco, Koch 75134 (ENCB); Tepotzotlán, 3 km al W
de Magi, Mendiola. 188 (ENCB); parte alta de la Sierra
1 de la estación de microondas, Rze-
dowski 29956 (CH. APA. E NCB); Texcoco, 20 km al NE de
Texcoco, Cruz 1834 (ENCB); Cerro Tetzahuitl, Chávez
2233 (ENCB): 24 km al E de Texcoco, García s.n. (CHA-
›А); 19 km al E de Texcoco, Koch 75134 (ENCB); Tlal-
Medellín s.n. (ENCB); Tlazala de Fabela, 4 km
Santiago Tlazala, Jiménez s.n. (ENCB); 7 km al E
CB); Cafiada de Onofre,
T
—
le Alcaparrosa, cerc
manalco,
al E de !
de Santiago Tlazala, Luna s.n.
Román 279 (ENCB); . Centro Ceremonial Toluca,
Cuellar s.n. (ENCB); Valle di Bravo, ladera E del Cerro
Gordo. — 9107 (ENC B); Villa del Carbón, Predio Pie-
dri a Azul, Ávila s.n. n NC B); San Jerónimo, Matuda 29014
„Matuda 29735 (ENC
г tahuacan,
; 5 km al SW de Cahuacán,
> Santiago Tlazala,
E de era Tisxala.
(E NC B): 2 km de Cahuacán, Mortecinos 200 (E
km al NW de Cahuacán, Reedow ski 33758 (ENCB).
de (
N
17. Quercus dysophylla Bentham, PI. Hartw.
55. 1840. TIPO: México. Huasca: Hartweg
421 (holotipo, K!).
132,
Esperanza,
Natl. Acad. Sci. 20:
TIPO: México.
Mem.
Quercus esperanzae Trel.,
pl. 248. 1924. Syn. nov.
Purpus 5332 (NY!).
Trel., Mem. Natl. Acad. Sci. 20: 177, pl.
Quercus — ri
35 )24. Syn. nov. TIPO: México. Cofre de —
— 255 (Р!).
Quercus — Trel., Mem. Natl. Acad. Sei. 20: 131, pl.
247. |. TIPO: México. Cajalpa, cerca de Tolue a,
к Fe (P).
Árbol de 5-20 m de alto,
diám., estriadas, glabrescentes, el indumento se ex-
ramillas de 1-3 mm
tiende a los pecíolos y nervadura central, formado
de tricomas estrellados estipitados, de color cas-
año-rojizo; lenticelas hasta de 1 mm de largo, más
claras que el tallo; yemas de 2-5 mm de largo,
ovoides, de color café claro, con el ápice redon-
deado y los bordes largamente ciliados; estípulas
caedizas, lineares, de 4-9 mm de largo, pubescen-
tes en toda su superficie; hojas maduras gruesas y
coriáceas, ovadas, lanceoladas o elípticas, lámina
(3235-15-17) X (12-56) ст, ápice obtuso o
agudo, mueronado o aristado, aristas de hasta 2 mm
—
de largo, base redondeada o cordada, a veces asi-
métrica, borde revoluto, entero u ondulado, a veces
con algunos dientes, con O a 13 aristas en cada
lado, de hasta 2 mm de largo; nervaduras de 9 a
15 en cada lado, rectas o algo curvas, se bifurcan
Annals of the
Missouri Botanical Garden
C
Figura 17.
en el tercio distal o cerca del margen; haz verde
pálido, glabro, con tricomas estrellados, estipitados,
dispersos sobre la lámina, abundantes sobre la ner-
vadura central, nervaduras impresas; envés con pu-
bescencia amarilla, formada por tricomas estrella-
dos, estipitados, y por escasos o abundantes
tricomas simples glandulares que dejan ver la epi-
dermis ampulosa, nervaduras elevadas; pecíolos de
(2—)4—7(—12) mm de largo. de 1-2 mm diám., den-
samente pubescentes, se ennegrecen con la edad;
flores desconocidas; fruto solitario o en grupos de
dos, sésiles o en pedtinculos de 1—2 mm de largo,
con lenticelas claras; cúpula turbinada, de 10—12
mm diám., de 8-12 mm de alto, borde recto, a
veces involuto; escamas no engrosadas en la base.
no adpresas, excepto las superiores, canescentes,
ápice obtuso, márgenes más obscuros; bellota ovo-
ide, pared interna de pericarpo lanosa, de 12-17
mm de largo, de 10—11 mm diám., de color castaño
claro, incluída la mitad de su largo en la ctipula.
Figura 17.
Quercus dysophylla.—A. Rama. —B. Fruto. —C. Tricomas. (Romero 1.)
Reconocimiento. Quercus dysophylla se recon-
oce por sus hojas lanceoladas o elípticas, con el
borde entero y aristado, envés con pubescencia
amarilla y epidermis ampulosa; puede confundirse
con Q. crassifolia y Q. hintonii. Quercus crassifolia
se reconoce por sus hojas obovadas, dentadas, ar-
istadas con el envés lanoso amarillo, anaranjado o
café, y epidermis ampulosa. Quercus hintonii se
distingue porque el envés de sus hojas posee epi-
dermis lisa y las ramas de los tricomas estrellados
son más largas que las especies anteriores.
Distribución y hábitat. México en los estados
de Distrito Federal, Guanajuato, Hidalgo, Estado
de México, Michoacán, San Luis Potosí, Tlaxcala y
Veracruz. Se le encuentra en bosques de Quercus,
Pinus—Quercus, bosque mesófilo de montaña y
bosque de Pinus perturbado, y se asocia con Quer-
cus rugosa y Q. castanea, en altitudes de 2490—
2850 m.
Fenología. Florece en marzo y abril y fructifica
en septiembre y octubre.
Volume 89, Number 4
2002
Romero et al. 583
Quercus en el Estado de México
Nombres populares y usos. Encino, laurelillo.
Bello y Labat (1987) sefialan que se utiliza como
lefia y postes de cercas.
MÉXIC O. Estado de Méxi-
Ejemplares examinados.
co: Amecameca, Alcansi,
Amecameca, Vela “i
de Chalma,
al S de Huixquilucan, Геб 713 (ЕМСВ);
alrededores de Huixquilucan, Román 332 (IZTA); Tepot-
zollán, Sierra de Alcaparrosa, Núñez 1677 (ENCB): parte
alta de la Sierra de Alcaparrosa, Rzedowski 31264
(ENCB); Timilpan, San Andrés — Camacho 357.
360, 241 (IZTA); Tlazala de Fabela, 4 km al E de Santiago
Tlazala, Jiménez 141 (ENCB): Villa del Carbón. Villa del
Carbón, Ramos 1 (IZTA); Villa Nicolás Romero, Villa Ni-
colás Romero, Jiménez s.n. (ENCB); 2 km al NW de Ca-
— ‘an, Rzedowski 33757 (ENCB); | km al NE de Ca-
асап, Tirado s.n. (ENCB)
. Quercus elliptica Née, Anales Ci. Nat. 3:
278. 1801. TIPO: México. Hidalgo: Ixmiquil-
pan y Zimapán, Née s.n. (I
Árbol de 8-15 m de alto, con el tronco de 15—
70 em diám., corteza gris obscura; ramillas den-
samente pubescentes, de 2-3 mm didm., al prin-
cipio de color amarillo, después se ennegrecen, el
indumento formado por tricomas estrellados unos
largos y otros cortos, erectos; lenticelas escasas cu-
biertas por la pubescencia, de hasta 0.5 mm de
largo: vemas de 2—4 mm de largo, ovoides, de color
castaño obscuro; estípulas de 7-8 mm, oblanceo-
ladas, ciliadas, principalmente en ápice v base.
membranosas, escariosas, decíduas; hojas jóvenes
con el haz v envés cubiertos de pequefios pelos
estrellados, los de la nervadura central más largos.
y pelos simples glandulares color ámbar: hojas
maduras rígidas, coriáceas, elípticas, lanceoladas u
oblanceoladas. lámina (2.1—)4.5-12 X (1.8—)3—5(—
6.5) em, ápice obtuso, agudo o emarginado, cuando
es agudo a veces con arista de hasta 2 mm de largo.
base redondeada, cordada o auriculada, borde en-
tero, engrosado, cartilaginoso, ligeramente revoluto:
nervaduras de 8 a 16 en cada lado, se ramifican en
el borde, forman arcos interconectados: haz verde
pálido, lustroso, casi glabro, con pequeños pelos
muy dispersos, en la base de la nervadura central
se agrupan pelos con rayos más largos, nervaduras
impresas o ligeramente elevadas; envés glabro, a
veces con pelos estrellados dispersos sobre la lám-
ina, con aproximadamente 6 rayos cortos extendi-
dos, en las axilas de las nervaduras son más largos.
nervaduras elevadas, pálidas, gruesas, epidermis
ligeramente ampulosa, papilosa; pecíolos de (3—)4—
7 mm de largo, de 1-1.5 mm diám., densamente
pubescentes; amentos masculinos de 5-9 cm de
largo, de muchas flores en posición alterna, flores
sésiles, perianto de 2—4 mm diám., de 1.5-2 mm
de largo, lóbulos ciliados, 4 a 7 estambres, anteras
oblongo-elipsoidales, de 1-1.5 mm de largo, gla-
bras, filamentos de 1-2 mm de largo, raquis muy
pubescente; flores femeninas solitarias, en pares о
en grupos de 4 a 5 sobre un pedúnculo pubescente
~
—
1-1.5 ст de largo: frutos de 1 a 3, sobre un
pedúnculo de 2-3 mm de largo; cúpula pateliforme,
de 3—4 X
no engrosadas en la base, ápice obtuso, pubescen-
14-15 mm diám., escamas triangulares.
te, bordes más obscuros: bellota ancha, pared
interna del pericarpo lanosa, ovoide de 13-15 X
14-15 mm didm., incluída en la cúpula menos de
una tercera parte. Figura 18.
Discusión. Quercus elliptica se reconoce por sus
hojas elípticas, ápice aristado y envés glabrescente.
Distribución y hábitat. En México en los esta-
dos de Guerrero, Jalisco, Estado de México, Mi-
choacán, Nayarit, Oaxaca, Sinaloa y Veracruz, tam-
bién en América Central. Habita en bosques de
Pinus—Quercus y de galería, en altitudes de 1500—
1900 m.
Fenología. Florece de febrero a marzo y fruc-
tifica en octubre.
Nombres populares y usos. Encino cucharita,
encino tapahuite, encino nanche, encino colorado,
encino laurel, encino cáscara. Su madera se usa
como leña y para elaborar carbón (González, 1986),
también para bancos, muebles rústicos, mangos,
cabos de herramienta, vigas de construcción y pos-
tes.
Ejemplares examinados. MEXICO. Estado de Méxi-
co: Tejupilco, Paso del Jate, Parque Cinegético Nanchi-
titla, Abundiz 525. 536 (IZTA); Reserva Cinegética de
Nanchititla, у=. 1966 (IZTA); en falda \ del cerro, Ma-
tuda 31544 (CODAGE d km 2 3 ELC ийни t ae hititla.
Rojas & Romero pes A); km 12 El Corupo-Nanchi-
titla, Rojas & Romero iir (IZTA); 2 I al NW de Nan-
chititla, Rzedowskt 22116 (ENCB); Nanchititla, alrede-
dores de la población, Rzedowski 30303, 30309 (ENCB).
19. Quercus hintonii E. F. Warburg, Bull. Misc.
Inform.: 91. 1939. TIPO: México.
México: Temascaltepec de González, Nanchi-
titla, Hinton 6359 (holotipo, K!).
Estado de
шс ig dg F. Warb.. Bull. Misc. Inform.: 95.
ПРО: México. Estado de México: Temascal-
— de González, Berros, Hinton 6568 (holotipo.
K!).
кн эга E. F. Warb., Bull. Misc. Inform.: 94.
. TIPO: México. Estado de México:
Р рес ae ;onzález, Mina de Agua, Hinton 6576 (hol-
otip
Temascal-
T—
Inform.: 90. 1939.
Quercus р F. Warb., Bull. Misc.
T Distrito de Te-
IPO: México. Estado de México:
Annals of the
Missouri Botanical Garden
Figura 18.
mascaltepec de González, Cuentla, Hinton 6577 (fo-
tograffa del holotipo, K!).
Arbol caducifolio, hasta de 15 m de alto; tronco
de 30-50 cm diám., corteza con placas cuadran-
gulares, en individuos muy jóvenes ésta es rugosa:
ramillas jóvenes de 1-4 mm diám., con abundante
pubescencia blanco-amarillenta, formada de trico-
mas estrellados, que se reduce y ennegrece con el
tiempo; lenticelas blancas, de 0.5-1 mm, más no-
torias en ramas viejas; yemas ovoides, de color cas-
taño obscuro, de 1-8 mm de largo, con escamas
coriáceas y pilosas; estípulas oblanceoladas, de 9—
13 mm de largo, membranosas, pilosas en márgenes
y base, decíduas; hojas jóvenes con abundante pu-
bescencia rojiza que cambia a amarillo en la mad-
Quercus elliptica.—A. Rama. —B. Inflorescencia masculina. С. Tricomas. (Romero 3991.)
urez; hojas maduras coriáceas, lanceoladas, ovado-
lanceoladas, obovadas o elípticas, lámina 5-21 X
3-10 cm, ápice aristado, corto a largamente acu-
minado, base obtusa, redondeada, cordada o aguda,
con frecuencia asimétrica, borde aristado, dentado,
a veces entero, revoluto, hasta con 7 aristas en cada
lado, es frecuente que éstas se presenten en un sólo
lado; las nervaduras se prolongan hasta las aristas,
las primeras de 6 a 14 en cada lado, rectas o lig-
eramente arqueadas, ascendentes, ramificandose en
el borde; haz verde claro, lustroso, glabro excepto
en la base y en la nervadura central, en hojas in-
maduras es rojizo a verde obscuro, con abundante
pubescencia formada por tricomas estrellados y
simples glandulares, con la madurez la pubescen-
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de Mexico
cia disminuye en cantidad y cambia de rojiza a
blanco-amarillenta; nervadura central a veces algo
elevada, nervaduras primarias y secundarias im-
resas; envés lanoso tomentoso, con tricomas es-
trellados de aproximadamente 1.5 mm de largo, con
las ramas entrelazadas, epidermis lisa, con abun-
dantes tricomas simples glandulares, de color ám-
bar, el tomento cambia de blanco en hojas jóvenes,
a amarillo en las maduras, nervaduras conspicua-
mente elevadas; pecíolo de 0.7-2.9 X 1-2 mm,
lanoso tomentoso, con frecuencia más ancho en la
base; amentos masculinos de 3-10 cm de largo con
muchas flores, raquis con abundante pubescencia,
perianto pubescente de 1-2 mm de largo y de 2 o
3 mm diám., amarillento, con frecuencia los bordes
~
son rojos, estambres 6 о 7, exertos, anteras oblon-
gas, glabras, apendiculadas, de 1-1.5 X 1 mm, con
€
frecuencia rojizas, filamentos hasta de 2 mm de
largo: amentos femeninos de 1—14 cm de largo, con
1 a 6 flores, raquis pubescente; perianto de 2 X
1.5 mm, amarillento: estigmas 3 o 4, espatulados,
de color rojo obscuro; frutos 1 a 4 en pedünculos
de 6—7.5 mm: ctipula poculiforme a pateliforme, de
10-14 mm diám., de 4—7
as con ápices de color castafio, pubescentes; bellota
mm de alto, escamas lax-
globosa a comprimida, pared interna del pericarpo
lanosa, de 6-15 mm de largo, de 9-13 mm Фат.
con el ápice plano a umbonado. Figura 19.
Discusión. Quercus hintonii se reconoce por sus
hojas generalmente lanceoladas, aristadas, con el
envés densamente tomentoso y epidermis lisa. Q.
hintonii muestra similitud con Q. crassifolia y Q.
dysophylla (ver descripción de Q. crassifolia).
Distribución y hábitat. En México en el Estado
de México. En bosques de Quercus—Pinus, se aso-
cia con Quercus magnoliifolia, Pinus oocarpa, P.
pringlei, Clethra mexicana y Juniperus flaccida, en
altitudes de 1300-1950 m
Fenología. Florece en marzo y los frutos mad-
uran en junio a agosto.
Nombres populares y usos. Encino prieto.
Su madera es empleada localmente para la ela-
boración de mangos de herramientas, vigas, postes
de cerca, bancos rústicos y leña. Podría utilizarse
para fabricar duelas de parquet, lambrín, muebles,
durmientes, pilotes para minas y barricas (Arcia,
— O. Estado de Méxi-
Matuda & col.
Ejemplares examinados.
со: Ата сегсапѓаѕ de alepec,
29823 (CODAGEM); Clachic — 7 km al NE, Pineda
1048 (INIF); 4 km al S de Amatepec, Pineda 1052 (INIF):
San Simón de Guerrero, km 83 Toluca-San Diego Cuentla,
Rojas, Romero & — 3218 ep ТИТА); Mina de
Agua, Rojas & Romero 3605 (ENCB,
Cuentla, Rojas & гич 3606 (ЕМС 3 a A); Sultepec,
>
تو
=
=
=
87
m
,
km 25 carretera a San Miguel кокк Torres 575
IZTA); Tejupilco, Cerro de Nanchi sarcía s.n.
(IZTA); эч ое le la poblac ión ds Tejupilco, -
NIF); 5 km al SW de Nanchititla, González
PH ub Pefia Bonita, González 5399, 5444,
5 (MEXU); Potrero Chicc — 36, 5438
(MEXU); Reserva Ecológica de Nanchititla, — 318
(IZTA); 5 km al NE de Palos Prietos, Pineda 4 (INTE);
Los Cuervos, Rojas & Romero 36 ENC P Te-
mascaltepec de Gonz — Chorrera, Hinton he (ENCB);
Pant ija, Hinton 6225 (ENCB); km 70 carretera Federal
Rojas & T 3400
de Temascaltepec de
jupilco, Rz Bead "E (IN
'anías, Matuda & ‹ sd
GEM).
ENCB, IZTA); 5 km
с^
—.
' 29827. 31588 (CODA-
20. Quercus laurina Humboldt & Bonpland, PI.
Aequinoct. 2: 32, pl. 80. 1809. TIPO: México.
Hidalgo: Cerro de las Navajas, cercanías de
Morán, Bonpland 4143 (isotipo, B!).
Quercus — is Benth., Pl. Hartw. 56. 1840. TIPO:
México. Hidalgo: Real del Monte, Anueg 2 ).
Quercus ‘ourgaet Oersted ex оаа Мет Acad.
Sei. 20: 2m pl. 366. 1924. TIPO: México. San Ni-
e de México, m он 1013
Quercus lanceolata Humb. & Bonpl., Pl. Reauinoot. 2: 34,
pl. 81. 1809. TIPO: México. Morán a Santa Rosa,
Bonpland s.n. (B).
Quercus caeruleocarpa, T em. Natl. Acad. Sci. 20:
163, pl. 321. 19 nd
México. Distrito Federal:
Contreras, 9 * Te Endlich 1365 a (B)
Árbol de 10-30 m de alto, tronco de 50 ст diám.
o más, corteza con grietas poco profundas y piezas
chicas, de color gris obscuro; ramillas de color gris,
de 1-2.5 cm diám., pubescentes al principio, des-
pués glabrescentes, pubescencia de tricomas es-
trellados pequefios; lenticelas menores de 0.5—3
mm de largo, de color claro u obscuro; yemas de
1.5—4 mm de largo, ovoides, agudas, de color cas-
tafio, escamas gruesas con el margen apical ciliado;
estípulas de 3—6 mm de largo, oblanceoladas o sub-
agudas, escariosas, membranosas, decíduas; hojas
jóvenes con pubescencia formada por tricomas es-
trellados pequefios y simples dispersos en haz у
envés, con el tiempo ésta se reduce a las nerva-
duras central y primarias; hojas maduras rígidas,
coriáceas, lustrosas, lanceoladas o elíptico-oblan-
lámina (2-)5-11(-15.5) X 1.5-4(-6.5
cm, ápice agudo o acuminado y por lo general ar-
ceoladas,
istado, base redondeada, cordada, atenuada o cor-
dada, borde entero o dentado, plano o revoluto, a
veces ondulado, engrosado, con 1 a 5 aristas de
cada lado, a veces se presentan sólo de un lado de
la hoja, distribuidas en el tercio superior; nerva-
duras primarias de 4 a 12 en cada lado, rectas о
ligeramente arqueadas, ascendentes, se ramifican
antes del margen; haz verde lustroso, todas las ner-
vaduras pálidas y elevadas, nervadura central pu-
586 Annals of the
Missouri Botanical Garden
Figura 19. Quercus hintonii.—A. Rama. —B. Fruto. С. Tricoma. —D. Hoja. —E. Flores masculinas. (Romero
3604.)
bescente en su base; envés lustroso, amarillo o a
veces más pálido que el haz, pubescencia de tri-
comas estrellados estipitados, restringida por k
general a las axilas de las nervaduras primarias,
donde se observan tricomas simples, blancos y es-
casos tricomas glandulares de color ámbar, epider-
mis papilosa, nervaduras elevadas; pecíolos de (2—)
25-15-22) mm de largo, de 0.5-1.5 mm diám., fin-
amente pubescentes, glabrescentes; amentos mas-
culinos de 3.54.5 em de largo, raquis negro con
Volume 89, Number 4
2002
Romero et al. 587
Quercus en el Estado de México
Figura 20. Quercus laurina.—A. Rama. —B. Fruto. (С.
—C: Torres 1468; D, E: Romero 3321.)
tricomas largos, perianto de 2-3 mm diám., con los
lóbulos largamente ciliados, anteras de 1.5 mm de
largo, filamentos de | mm de largo: fruto anual o
bianual, solitario, en pares o en grupos de 3, sésiles
o en pedúnculos de 3-12 mm de largo: cúpula
hemisférica. de 10—15 mm diám., escamas leñosas,
Tricomas. —D. Hoja. —E. Inflorescencias masculinas.
no engrosadas en la base, pubescentes, de ápice
obtuso: bellota ovoide, pared del pericarpo lanosa,
de 7-19 mm de largo, de 7-12 mm diám., incluída
un tercio de su largo en la cúpula. Figura 20.
Discusión. Quercus laurina se reconoce por sus
588
Annals of the
Missouri Botanical Garden
hojas aristadas, el envés con pubescencia restrin-
gida a las axilas de las nervaduras. Esta especie
posee similitud morfológica con Q. affinis y se dis-
tingue por poseer yemas florales de forma conoidal,
base de las hojas cuneada y nervaduras planas;
mientras que Q. laurina posee las yemas ovoides,
base de las — atenuada o redondeada y nerva-
duras elevadas.
Distribución y hábitat. En México en los esta-
dos de Distrito Federal, Guanajuato, Guerrero, Hi-
dalgo, Jalisco, Estado de México, Michoacán, Mo-
Querétaro,
Zacatecas. En bosques de Pinus—Quercus,
relos, Puebla, Oaxaca, Tlaxcala y
Pinus—
Abies, bosque mesófilo de montafia y encinares per-
turbados, en altitudes de 2240-3150 m.
Fenología. | Florece en abril y fructifica de junio
a octubre.
Nombres populares y usos. Chilillo, encino co-
lorado, atlualpitzahul, encino blanco, encino lau-
relillo, encino roble, encino xicatahua, tesmolera,
encino hoja angosta, huitzalacate, encino prieto,
encino uricua, encino chilillo.
Se utiliza para bancos, muebles rásticos, cabos
de herramienta, vigas de construcción, papel Kraft
y fabricación de chapa (González, 1986) y como
lefia.
Ejemplares examinados. MEXICO. Estado bee Méxi-
co: Acambay, cerro Detiña, шо 1468 (IZTA); Muy-
teje, Quintero & Rojas 18, 22 (IZTA); Ar ne de Be-
cerra, 5 km al NW LARUM Be ndita, Pineda 1347 (INIF);
4 `сатеса, barran Antonio — Ern
159 (ENCB); Alcansi, y & Mancera s.n. (ENCB); 1 km
al E de San Antonio Zoyatzingo, ا 2643 32
(ENCB); 1 km al E de San Antonio, base del Ixtaccihuatl,
Xelhuantzi 5081 (I F); — cumbre del Cerro Chil-
иса, Rzedowski 32611 (ENCB); Atlautla, Tecomazusco,
camino de Atlautla, al paraje de s Santa Teresa, Herndndez
40 (INIF); Capulhuac, Cerro l km al W de San
Miguel Almaya, Rodríguez s.n. (ЕЛ NC SB 6
al 5 de Coacalco, vertiente N de la
е 30768 (ЕМСВ); Sierra det Guadalupe, €
d ~ alco, Rzedowski 30774 (ENC
15 — al i
rimas, geste s.n.
Coatepec, Опе ra 42 (ENCB); 5 ES al E de Coatepec,
Román 315 (ENC * Chalco, cerca de la colonia Agrícola
M: vila Camacho, Cabrera s.n. (ENCB); Chicoloa-
pan, 5 km al E de Coatepec, — 311 (ENCB); El Oro
de Hidalgo, La Cima, Pérez 19 (IZTA); La Cima, Valdés 1
(1Z i : Sd Matuda 28101 (ENCB);
Ixtapaluca, Estación Experimental de Enseñanza e Inves-
+ ión de Zoquiapan, е 1160 (ENCB); 1 km al
» de Llano Grande, Rivas & Campos 13 (ENCB); 6 km
2s SW de Río Frío, Roe & Scott. 1440 (ENCB); Ji-
quipilco, vertiente W de la Sierra de las Cruces, Alvarez
8, 15 (IZTA); km 15 carre : ra Jiquipilco-Villa del Carbón,
Romero & por 3176 (IZTA); Jilotepec, San Bartolo, Ma-
tuda & col. 29063 (ENC ^s Santa Ana — 3kma
E de San — cen in, Fernández 36 (ENCB); Lerma,
Santa María Tlalmimilolpan, cerca de Lerma, Franco 63
|
7
1
erra * Guadalupe.
era
m
2°
(ENCB); cerca de
calpan de — Villa
Guarda, Herr 19]
Salazar, Villanueva s.n. (ENCB); Nau-
Alpina, Corona s.n. (ENCB); El
(ENCB); Santiago Tepatlaxco,
Jiménez 132 (IZT A); Los Remedios, Maarigal 74 (INIF);
San Francisco Chimalpa, Ruiz 33 (1 Ocuilan de Ar-
teaga de Arteaga, km 18 terracería Ocuilan de Arteaga—
Cuernavaca, Castañeda, Lugo & Trejo 13 (VZTA); San Mar-
tín de las Pirámides, Cumbre del Cerro Gordo, Rzedowski
18810 (ENCB); Sultepec, La Cieneguilla, — 163
(ENCB); San Miguel йш э y Torres 3 302 (IZTA); Te-
1 & Tejero s.n. (ENCB)
zález, un al de X PASE
IZTA); Comunal de Tequisquia-
pan, Huerta T-8 (ENCB); Teotihuacán, Cerro Ge јат al
de Teotihuac ‘in, Espinosa 635a (ENCB); Texcoco, 24 km
al E de Texcoco, sobre brecha rumbo a Tlaloc, García s.n.
(ENCB); 8 * al SE T Tequesquinahuac, Perino 3137
(ENCB); Сай: ada al 5
masc —
tura 1123 (INIF); Timilpan, — de San Andrés, Ca-
macho 248 (IZTA); Tlaln 1analco,
(INIF); 4 km de San Rafael, Magafia 1057
(NIF): a Medellín 227 (INIF); 3 km al S de
San Rafael, Ramá ENCB); Villa del Carbón, El
Plan, González s.n. ( Т А), El Gato San Jerónimo, Rojas
«€ Romero 3168 (IZTA); Villa Nicolás Romero, Transfi-
"uración, Ocaña s.n. (УТА); Zacualpan, cerro La Corona,
Машаа 30324 (CODAGEM).
21. Quercus mexicana Humboldt 9 Bonpland,
Pl. Aequinoct. I: 35, pl. 82. 1809. TIPO:
México. Guanajuato: Santa а Bonpland
4218 (B)
— rugulosa Mart. & Gal., Bull. Acad. Roy. Sei.
— Vol. 10, pt. П: 209. 1843. TIPO: México.
edro y San Pablo, cerca de Real del Monte,
XM 116 (BR
—
Árbol de 3-15 m de alto, corteza gris; ramillas
de 1.5-2 mm diám., al principio con pubescencia
formada de tricomas estrellados, pronto glabrae,
corteza gris, con lenticelas menores de 0.5-2(-3)
mm de largo, casi del mismo color que la corteza;
yemas de (1-)2-3(-6) mm de largo, ovoides, ápice
agudo, de color castaño, escamas pubescentes en
sus márgenes; estípulas de 3-5 mm de largo, es-
cariosas, lineares, caedizas, algunas persistentes
por un tiempo cerca de las yemas apicales; hojas
jóvenes del mismo color que las adultas; haz con
el indumento formado de tricomas estrellados muy
pequefios dispersos, envés con tricomas estrellados
dispersos; hojas maduras decíduas, elípticas, lan-
ceoladas u oblongas, coriáceas, lámina (2-)3-9(-
12) X (0.9—)1.5—3.5(—4.2) em, ápice agudo, suba-
gudo o redondeado, con una arista de hasta 2.5 mm
de largo, base ligeramente revoluto, engrosado, car-
tilaginoso, borde entero; nervaduras de 6 a 12 en
cada lado, ascendentes, casi rectas, se bifurcan en
el ápice; haz verde obscuro, con algunos tricomas
estrellados dispersos en la base de la hoja y en la
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
SAN S
ESTA
—
[A es.
RAH 9
B
sp.
ek
A
Figura 21. Quercus mexicana.—A. Rama. —B. Fruto.
nervadura central, ésta impresa o ligeramente ele-
vada, las secundarias impresas; envés ligeramente
más pálido, con pubescencia de tricomas estrella-
dos contortos, que se observan como puntuaciones,
uniformemente distribuidas, epidermis papilosa:
pecíolos pubescentes, de (2-)3-8 mm de largo, de
0.5-1.5 mm diám.; flores desconocidas; fruto anual,
solitario o en pares, sésiles o en pedúnculos de 2—
9 mm de largo; cúpula hemisférica, de 10-13 mm
diám., con las escamas delgadas, no engrosadas en
la base, pubescentes, con el ápice subagudo o re-
dondeado, márgenes por lo general glabros: bellota
ovoide, pared interna del pericarpo lanosa, de 9—
15 mm de largo, de 8-11 mm diám., incluida un
tercio o la mitad de su largo en la cúpula. Figu-
ra 21.
Discusión. (Quercus mexicana se reconoce
porque sus hojas tienen bordes enteros, ápice ar-
istado y envés con tricomas contortos que se ob-
— C. Tricoma. (Quintero 30.)
servan como puntuaciones. Quercus mexicana
muestra similitud con Q. crassipes, esta ültima se
distingue por no presentar los tricomas contortos a
manera de puntuaciones.
Distribución y hábitat. En México en los esta-
dos de Chiapas, Distrito Federal Guanajuato, Hi-
dalgo. Estado de México, Nuevo León, Puebla,
Querétaro, San Luis Potosí, Tamaulipas y Veracruz.
En bosques de Quercus y Quercus—Pinus, también
en suelos erosionados y a orillas de arroyos, se aso-
cia con Pinus montezumae, P. leiophylla, P. teocote
y P. rudis, en altitudes de 2230—3050 m.
Fenología. | Fructifica de julio a enero.
Nombres populares y usos. Encino, encino te-
zahuatl.
Hasta el momento no se tiene conocimiento so-
bre sus usos.
Fjemplares examinados. MÉXICO. Estado de Méxi-
co: Atizapán, 4 km al N de Atizapán, Castilla 2725
Annals of the
Missouri Botanical Garden
(IZTA); Coyotepec, Sierra de Alcaparrosa, Rojas & Romero
156 (IZTA); о Fraccionamiento La Herradu-
ra Viel. ER s.n. ОМІР); Jilote pec. Ripe Matuda
26742 (CODAGEM); Las Peñas, Rojas € Romero 3076
(ZTA): Ojo de Agua, Rojas & Romero 2 — =
mascalapa, Sierra de Guadalupe Р. 3 Padres, Ortíz s
(IZTA); Cerro Gordo, Tejero & Castilla 634 zT А); Te.
potzotlán, Arcos del Sitio, Ceballos 2 (IZTA); Villa del
Carbón, cortina de la — San Luis Taximay, Tejero &
Abundiz 2766 (ZTA): S in Luis Taximay, Tejero & Castilla
2723 (IZTA); Villa Nicolás Re m | km al E de Lan-
zarote, Rzedowski 31877 (INIF); З km al NW de Cahu-
асап, Rzedowski 32628 (INIF).
22. Quercus seytophylla Liebmann, Overs.
Danske Vidensk. Selsk. Forh. Med-
180. 1854. TIPO: México.
Oaxaca: Yalala a Yagochi, 144—
6 =3557 (C)
Kongel.
lemmers Arbeider:
Liebmann
Árbol de 8-20 m de alto, de 30-50 em diám.:
ramillas de 1.5-3 mm diám., grises a castaño rojizo
o negras, finamente pubescentes, con tricomas pe-
queños estrellados; lenticelas pálidas, de 0.5-3 mm
de largo; yemas de 1-4 mm de largo, ovoides o
elípticas, escamas pubescentes con los márgenes
ciliados, engrosadas en la base, de color castaño:
estípulas lineares, pubescentes, caedizas, membra-
nosas, de hasta 5 mm de largo; hojas jóvenes pu-
bescentes, haz verde amarillento con tonos cafés,
con numerosos tricomas pequeños, estrellados, se-
mejantes a los que se observan en las ramillas, tri-
comas de mayor tamaño se concentran en la ner-
vadura central; hojas | maduras — coriáceas,
generalmente obovadas, elípticas, oblanceoladas o
lanceoladas, lámina (6.2—)8-14.5(-19.5) х (2.3-)
4—7(—8.5) em, ápice con un diente alargado, aris-
tado, base obtusa o cuneada, a veces oblicua, borde
engrosado, revoluto, cartilaginoso, con Та 7 dientes
aristados a cada lado, casi desde la base de la hoja,
aristas hasta de 4 mm de largo; nervaduras 5 a 9
primarias en cada lado, rectas о arqueadas, anas-
tomosándose cerca del borde. pasando directa-
mente al diente y luego a la arista, a. veces las
divisiones de las primeras lo hacen también; haz
verde grisáceo, no lustroso, glabro, excepto en la
base de la nervadura central, generalmente los tri-
comas ennegrecidos, nervaduras impresas, el as-
pecto general es rugoso; envés densamente pubes-
cente, blanquecino, que cambia a amarillento y que
se ennegrece con el tiempo, indumento formado por
tricomas estrellados con más de 10 rayos, sésiles,
epidermis ampulosa y papilosa; nervaduras central
y secundarias conspicuamente elevadas, blancas;
pecíolos de (0.7—)1—4 em de largo, de 1-2 mm
diám., se engrosan hacia la base, con fina pubes-
cencia gris; amentos masculinos laxos, perianto de
1.5-2 mm diám.,
rlabros, anteras de 1.5 mm de
B
largo, filamentos de 1.5-2 mm de largo: fruto anual,
en grupos de dos, sobre pedünculos de 5-15 mm
de largo; cúpula hemisférica, de 10-12 mm diám.,
escamas delgadas, pubescentes, largamente cilia-
das en los bordes, ápices redondeados a agudos,
adpresas; bellota ovoide, de 8-10 mm diám., de 10
mm de largo, incluída de un medio a dos tercios
de su largo en la cúpula. Figura 22.
Discusión. Quercus scytophylla se reconoce por
sus hojas dentado-aristadas, haz opaco y envés con
pubescencia blanca.
Distribución y hábitat. En México en los esta-
dos de Durango, Guerrero, Jalisco, Estado de Méx-
ico, Michoacán, Nayarit, Oaxaca, Puebla, Sinaloa y
Sonora. En bosque de Quercus, bosque de Pinus—
Quercus y bosque mesófilo de montaña, se asocia a
Arbutus, Pinus oocarpa, Pinus teocote, en altitudes
de 900-2500 m
Fenología. Florece en febrero y fructifica en
octubre.
Nombres populares y usos. Encino prieto, en-
cino blanco, encino rosillo.
Su madera se utiliza para la elaboración de tra-
bajos artesanales, así como fuente celulósica de
pulpa para papel (González, 1986).
MÉXICO. Estado de Méxi-
co: Amatepec, entre Amatepec y Sultepec, Matuda & col.
29832 (CODAGEM); Nuevo Santo Toni: di los Plátanos,
— E col. 29387 (ENCB); Sultepec, El PAREN:
POINTE examinados.
erca de Sultepec, Madrigal 17 (INIF); Las Trojas,
ríguez 179 (МЕ); Sulte — Rodriguez 181 (ENCB):
Tejupilco, dea Chico, 15 km a ' de Nanchititla,
González et al. 5436 (E NCB): Рейа Bonita, Nanchititla,
Ramos s.n. a NC B); 2 km al NW de Nanchititla, Rze-
dowski 2 7 (ENCB); Temascaltepec de González, km 65
le ( — Toluca-Temascaltepec, Casillas
(IZTA); Real de Arriba, Huerta, nos
Rar & col. RA-6
(ENCB, INIF); presa La Carbonera, Nah 3652 (ENCB,
INI); km 30 de la — ra federal 130, cercanfas de
‘ ZTA); Real de Arriba, Ramos
«€ Rocha 96 (INIF); n 51 de la carretera Toluca-San
Diego Cue tla, Rojas & МА 3212 (INIF); Valle de
Bravo, Huerta & col. CV-23 (ENCB, INIF);
Valle de Bravo, aeg 2604 de NCB); Zacualpan, km 6
de la carretera Zacualp: p е 152 (IZTA):
Lac — Маш & are 30298, 30299 (E
3 de la carretera Campana de Plata-Subestación C oronas,
Timo & Castilla 1734 (IZ TA).
Casas Viejas,
eer
23. Quercus urbanii Trelease, Proc. Amer. Phi-
los. Soc. 60: 32, pl. 2. 1921. TIPO: México.
Michoacán a Guerrero, Sierra Madre, 20 June
1899, E. Langlassé 1066 (B).
Árbol de 4—10 m de alto, tronco de 20-30 em
diám.; ramillas gruesas, con costillas, de 5-11 mm
diám., densamente pubescentes, pubescencia
amarillenta, gris-amarillenta o negra, formada por
Volume 89, Number 4
2002
Romero et al.
Quercus en el Estado de México
Figura 22.
tricomas estrellados estipitados con las ramas er-
ectas y por abundantes tricomas glandulares sobre
la epidermis, corteza castaño-rojiza a negra; len-
ticelas inconspicuas debido a la pubescencia, en
las ramas таз gruesas éstas miden hasta 4 mm de
largo; yemas de 5—10 mm de largo, ovoides, color
castaño, las escamas engrosadas en la base, gla-
brescentes las exteriores, densamente pubescen-
tes las interiores; estípulas de 7 mm de largo, de
| mm diám., pubescentes principalmente en la
base y bordes, persistentes en las yemas; hojas
jóvenes pubescentes, haz cubierto por tricomas
glandulares rojizos, tricomas simples dispersos y
tricomas estrellados largos, estos últimos princi-
palmente en los bordes, envés con pubescencia
blanca formada por tricomas estrellados muy lar-
gos: hojas maduras, rígidas, gruesas, pandurifor-
mes, obovadas, suborbiculares, orbiculares u ova-
Quercus scytophylla.—A. Rama. —B. Fruto.
—JC. Tricomas. (Orozco 49b.)
do-elípticas, lámina (12-)15-30 X (12-)17-34
cm, a veces más larga que ancha, ápice obtuso, a
veces algo escotado, base profundamente cordada,
borde revoluto, grueso, cartilaginoso, entero, den-
tado u ondulado, con (5—
—
10-20 aristas por lado,
en las % partes superiores, a veces desde la base,
aristas de hasta 4 mm de largo: nervaduras de 9
a 11 en cada lado, ascendentes, algunas forman
directamente las aristas, la mayoría se ramifican
cerca del borde; haz verde pálido, algo lustroso,
rugoso, glabro, excepto en las nervaduras princi-
pales y primarias en donde se encuentran tricomas
glandulares simples y estrellados, nervaduras im-
presas a ligeramente elevadas; envés con tomento
de aspecto lanoso, amarillento, formado por tri-
comas estrellados, estipitados con rayos de hasta
3 mm de largo, enredados, cubriendo uniforme-
mente la epidermis papilosa y glandulosa, nerva-
592
Annals of the
Missouri Botanical Garden
Figura 23.
duras elevadas; pecfolos de 3—4 cm de largo, de
2—4(—6) mm diám., tomentosos, se ennegrecen con
la edad; amentos masculinos de 13-23 cm de lar-
go, raquis densamente tomentoso, perianto de 6
mm diám., glabro, excepto los bordes de los lób-
ulos que son ciliados, anteras apendiculares, de
2-2.5 mm de largo, filamentos de 2 mm de largo;
amentos femeninos de 10 a 20 flores en pedün-
culos gruesos, con tomento abundante de color
amarillo; frutos en grupos de 5 a 10 o más, en
Quercus urbanii.—A. Rama. —B. Inflorescencias masculinas. С. Tricoma. (Torres 197; González 5017.)
pedünculos de 7.5-9 cm de largo, de 4-5 mm
diám., con denso tomento amarillento que se ob-
scurece y cae con el tiempo; cúpula de 8-12 mm
diám., escamas pubescentes con los ápices redon-
deados a truncados; bellota ovoide, pared interna
del pericarpo lanosa, de 10 mm de largo, de 8—
10 mm diám., incluída de un tercio a un medio
de su largo en la cápula. Figura 23.
Discusión. Se reconoce por sus hojas grandes,
Volume 89, Number 4
2002
Romero eta 593
Quercus en К Estado de México
acucharadas, el tomento del envés amarillo y sus
ramillas gruesas
Distribución y hábitat. Еп México en los estados
de Durango, Guerrero, Jalisco, Estado de México,
Nayarit, Sonora y Zacatecas. En bosques de Pinus—
Quercus y bosque mesófilo de montafia, asociado con
Quercus scytophylla, Q. crassifolia, Q. obtusata, Q.
laurina, Pinus montezumae, pseudostrobus, P.
oocarpa y Alnus, en altitudes de 1400-2500 m.
Fenología. Florece en diciembre y fructifica en
octubre.
Nombres populares y usos. Encino.
Su madera se usa como leña (González, 1986);
se sabe que también es utilizada para la elabora-
ción de arados y los frutos sirven para alimentar
ganado porcino (Vázquez, 1992).
— examinados. MEXICO. _Estado de Méx
co: Sultepec, entre Sultepec y el paraje El C E
Madrigal 25 (ENCB, INIF); La Cieneguilla Rodríguez
I (INIF); km = carretera Sultepec-San Miguel —
polora Torres 197 (IZTA); km 18 carretera Sultey
коо. Torres 428 ; Tejupilco, 5 5 km al
Sw ER Nanchititla, — 5017 (ENCB); ea hormi-
gueros, Nanchititla, ández 1747 (INIF); 12 km de El
Corupo a Nanc t эң "Rojas & Romero 3955 EZTA):
Puerto del Embocadero, 7 km al W de Luvianos, Rze-
dowski 22126 (ENCB, INIF); Nanchititla, alrededores de
a Población, Rzedowski 30307 (ENCB); Zacualpan, Zac-
ualpan, Matuda 30314 (CODAGEM).
Literatura Citada
Aguilar, E. & S. Romero. 1995.
cuatro especies de encino e descritas por War-
31:
Estudio taxonómico de
. 1979. Anatomía y RUM Físicas de la
ra de Tres = m e En ncino del Estado de Méx-
» Tesis. Esc. Nac a
M. & J. aa. 1987.
Estado de Michoacán, México. Centre d'E s
caines et жин ре Instituto Nacional de In-
vestigaciones Mia ales, México.
. Estudio del Uso del Bosque para la
Extracción d i Madera para Construcción de Ca-
sas y Fabricación de Herramientas en una Comunidad
Otomí. San Andrés o = de México. Tesis
de Licenciatura, UNAM, о.
Los Encinos — del
Mexi-
— A. 1938. Les chénes, monographie du genre Quer-
s. Paul Lechevalier, Paris.
Chin no, V. & Р. Jacques. 1986. Contribución al Conoci-
miento de la Flora Medicinal de (denen, Puebla.
Tesis de Lic ii iatura,
De La Paz O., C. 1976. Características Anatómicas de
Cinco Fins de México. Bol. Téen. Inst. Nac. Invest
Forest. No. 46. México.
19 82. — tructura Anatomica de Cinco Especies
del Gina ro Quercus. Bol. Técn. Inst. Nac. Invest. For-
esl., México, No. 88.
. 1985. Características Anatómicas de Siete Es-
к: ies del Сёпего Quercus. Bol. Тёсп. Inst. Nac. Invest.
. México, No.
— V. 1986. Сони ion al Conocimiento del Gé-
nero — (Fagaceae) en el Estado de Jalisco. Inst.
Bot. Univ. Aut., Guadalajara, Méx xico.
INEGI. 1987. Sintesis Geográfica. Nomenclator y Anexo
Cartográfico del Estado de México. Secretaria de Pro-
gramación y Aag a sto, México.
Martínez, M. 1954. Los Encinos del Estado de México.
Conia ba — Exploradora del Estado de México.
Gobierno del Estado de México, Dirección de Agric 'ul-
tura y Ganadería, Toluca, Estado de México, México.
1974. Flora Novo-Galiciana. Contr. Univ.
2
. & R. Mc Vaugh. 19 1972. The oaks (Quercus) de-
sc led by Née (1801), and {е аА & Bonpland
(1809), with comments on related species. Contr. Univ.
522.
c classification of —
Ann.
934. Infrageneri
(Fagaceae) p b rH ation. of sectional. names
Sci. Forest
| — genus (шеси m in Mexico. Pp. 447—
458 en T. P. Ramamoorthy, . Lot & J. F
EE Biological Diversity a Mexic ‘o: Origins and
Distribution. Oxford Univ. Press, Oxford.
The Tepehuán of —
Utah Press, Salt La
e 1. W. 1969.
eir Material Culture. Univ.
City.
Romero, R., C. s & S. Gómez. 2000. Flores hermaf-
roditas de — glaucoides Mart. & Gal. (Fagaceae)
= el Estado de Michoacán, México. Acta Bot. Mex. 52:
9—54.
— J. 1978. Vegetación de México. Ed. Limusa,
éxico.
e Rzedowski. 1979. Flora ded iy
ү” p Méxic o 1: 104-114. Ed. Continental S
Teese Ww. 1924. The American oaks. Mem. Natl. Acad.
. 20: 1-55, pl. 1-420.
El Género Quercus т en el
Tesis, UNAM, México
ушш. ү. 1993
Estado de Puebla:
ANNALS OF THE MISSOURI BOTANICAL GARDEN: CHECKLIST FOR AUTHORS
SCOPE
The Annals publishes original articles in systematic
botany and related fields. Papers whose purpose is the
establishment of new nomenclatural
plants and bry Rather, they
should be submitted to Novon for consideration. (Noron
phytes are not accepted.
instructions are available on the Web at www.mobot.org or
from the managing editor.)
CONDITIONS FOR PUBLICATION
Authors are urged to contact the — editor prior
to —— manuscripts. Because the dis
file will be used by the printer for мерна it is im-
portant to des to the items listed in the format section
of the checklist. Authors’ electronic files are formatted by
the editors and are sent t
tte/electronic
the printer in electronic form
(not as camera-ready page
Manuscripts that have ge n reviewed before submission
will be subject to the full normal review process initiated
here. Manusc ‘ripts not properly — may be ише!
for revision prior to review. It is the author's responsibility
missing ашын SA nen
su the editor
does not pe огт library researc
to
е
=
pt by prior arrange-
ment with authors not having access to crucial reference
materials. Communication with the editor at all points of
the publication process is encouraged.
PAGE CHARGES
Page costs are $80 per page. although charges may
sometimes be reduced. Changes in proof made by authors
will be billed non-negotiably to the author at $3.50 per
line changed.
CORRES
TUNNE,
Amy McPherson, Managing Editor
* Press, P.O. Box 299, St.
US
а — MBG Press, Monsanto Center,
Blvd., St. Louis, MO 63110-229
Tel: (314) 577-5112; Fax: (314) 577-9594.
Email: annals@mobot.org
Louis, MO 63166-0299,
4500 Shaw
www.mobot.org
1. General
[] Text is in English or Spanish on numbered pages.
LJ Manuscript is submitted on MS-DOS/WINDOWS 3%
in. diskette, as Microsoft Word? [preferable] file.
L] Three copies of double-spaced priis d manuscript, in-
cluding abstract, legends, tables, bs imen lists, Li
erature Cited, and footnotes, are enclosed. The printed
manuscript is the same as the — on the dis-
elte.
[|] Photocopies of the figures are included with each copy
of the manuscript. Copies of half-tone figures (photo-
graphs) and plates must be of sufficiently good quality
to show all the detail of the originals. Digital or dye-
sublimation proofs of electronic graphic files are sub-
entities in vascular
O0 0 Û Û
OO
O
OT
mitted. [Send originals or photostats of originals when
manuscript is accepted for publication.
Authorities, abbreviated according to Brummit & Pow-
ell's Authors of Plant Names, are pro
names the f
vided for species’
t time they are mentioned in the text;
OR they are provided in a table that includes names
of species.
Letter is enclosed containing the author's full mailing
address, telephone number, fax number, and e-mail
address
2. Format of Text
Right-hand — is not justified, and words are not
hyphenated ther
One space is ins — after periods, colons, question
=
marks, and all other punctuation mar
First hs аге indented with a tab.
Iwo hyphens with no space between them are used to
indicate ranges of n
bers
2500 m.
A single hyphen is used in hyphenated words, for ex-
ample, 9-nerved, oval-triangular.
t sentences of paragrap
umbers, for example, page num-
; measurements, 5—7 mm; altitudes, 1500—
Common Latin words or phrases are not italicized (e.g.
et al., i.e., sensu, etc.
Only taxon names at the rank of genus and below are
italicized.
Caria accents, umlauts, and other diacritical marks
have been included.
All figures and tables are cited in the text and are
numbered in the order in which they are to appear.
3. Style
Recent issue of the Annals is used as a model.
Chicago Manual of Style, latest edition, is used as a
reference
4. First Page
Footnotes are typed as double-spaced paragraphs on
the first page. The first footnote contains a
ments, including information on granting agenci
herbaria that loaned specimens, and the name of he
. The second footnote is the author's address. Ad-
— for additional authors are each separate foot-
notes. No other footnotes occur, except in tables, where
unavoidable.
artist
Abstract & Key Words
A one-paragraph abstract precedes text. Papers in
Spanish have an English abstract in addition} to a
Spanish resumen.
The abstract is concise (1 paragraph) and includes
brief statements about the paper's intent, materials and
methods, results, and significance of findings.
A brief list of ke y words immediately follows the Ab-
stract.
Volume 89, Number 4
2002
595
[]
O
O
O
O
O
O
O
O
O
O
Taxonomic Treatment
Species entries are organized as follows: Heading, Lat-
in diagnosis (if necessary), description, distribution,
vernacular name(s), discussion, specimens examined.
The discussions are parallel and follow the same order,
e.g., diagnostic characteristics, distinction from similar
species, variation, distribution and ecology, nomencla-
ture and typification, uses.
One paragraph per basionym is used as follows: Taxon
author, — citation, Spa citation, e.g ch-
miedia latifol — s) Sa. Nishida, < Mistoun
Bot. Gard. 86: 68 van Joni hil Nees,
Syst. Laur.: 674. hu E: Peru. Locality not in-
1835 (fl), dar 1433 — Е!;
еп).
dicated, iso-
ypes, BM!, E!, K!, LE not seen, OXF not se
Synonyms based on different — are placed i in sep-
arate paragraphs, each beginning with
followed by other combinations (if appropriate). and
the basionym,
citation of the
A brief Latin — for each new taxon is provided
rather than a complete Latin description.
s with infraspecific taxa: D ae and
discussion are composite (incorporating all infraspe-
For species wil
cific taxa) and parallel with other species desc a
Descriptions of и taxa are parallel with one
another (in the same species). All — are listed
under the ade infraspecific taxc
Desc poe Descriptions are lle sin a given
ank
—
easurements are metric. Hyp
ade a
1 . Hyphens are used
ir pare do extreme 2-)14.:
8.0(-31.9) cm long,” ee values are
not expected: ovary with (2)4(6) loc * Length X
: lamina 36.4—
intermediat
— are given in the following manner
J.1-16
W he n Go ant, nomina nuda, misapplied names, and
superfluous names are included in the discussion fol-
lowing the description, but are not part of the formal
synonymy.
Citation of Types
Exclamation points are used for specimens examined,
and types not seen are indicated as such (e.g., !
US not seen
Lectotype за, are included together with an in-
dication of where and the
thor. This arate is listed in
the ie of the paper inue is making the lectotyp-
‚ were — the year,
e Literature Cited. If
ification, the phrase “here designated” is used.
Tables
Tables are neat, double-spaced, and clearly presented.
In most cases the printer will typeset these from hard
Captions are — double-spaced as paragraphs at the
tops of the tables.
Each table starts on a separate sheet.
O
O
O
O
O
O
9. Abbreviations
Periods are used after all abbreviations (which are
minimized) except metric measures, compass direc-
and herbarium designations.
When dates are given as part of collection information,
three-letter month abbreviations are used, t for
months with four letters, which are TE out in full.
States are not abbreviated, and cities are spelled out.
[St.,
lions,
ouis, is acceptable. |
Periodicals are abbreviated e to B-P-H (Bo-
tanico-Periodicum- Renan and to B-P-H/S (Bo-
— -Periodicum- = инат Хирот
eviated according to Brummit
: M рне мн of Plant Nan
Book titles are abbreviated acc dis to Taxonomic ES
erature, edition 2, but with initial letters capitalize
ook t itles are spelled out in the Literature Cited.
Ifa m does not appear in B-P-H or TL-2, or if
Th uo are not available, its title is fully
spelled out
Herbaria are abbreviated according to Index Herba-
riorum, edition eight (8).
Abbreviated forms are not used for references in the
text, except when citing the names of plants. If it
necessary to cite a particular page in the text, the form
Smith (1998: 12) i
7
s usec
10. Specimens Examined
If many specimens were examined, those cited in the
text are limited to ca. 1% manuscript pages
index to specimens
following
If there are a large number, an
examined is placed at the end of the paper,
arra
`
the Literature Cited. It is
collector, followed by een number, follow
the number of the taxon in the text. Names (including
initial(s)) of first and second collector are provided, “et
al.” if three or more.
Specimens are cited in the text as follows: Additional
specimens examined (or Selected specimens examine
MEXICO xaca: Sierra San Pedro Nolesco, Talca.
12°37'N, 85° 14W. 950-1100 m, 3 Feb. 1987 (fl), Jer-
gensen 865 (BM, G, K, US). [Dates and reproductive
status are optional but are omitted from longer lists. |
aged alphabetically by
ed by
Countries are run together in the same dide e.g..
COU E A RY A. Major political division: ... COU N-
TRY 1 Major political division: . — par-
асга — are used for major continental regions within
major political divisions.
Specimen Vouchers and Genetic Sequences
If paper presents original data, associated herbarium
vouchers are cited. [ Vouchers for seed and/or other col-
lections should be included — зр Depen-
dent on the paper, reference to the inal wild source
may be required.| Vouc ‘hes are Шы cited for common
names and uses taken from specimen labels
Herbarium vouchers state the collector and number,
herbarium in which the voucher is located, and a clear
annotation that the material represents the voucher for
the study in question
[] Nucleic acid or protein sequences corresponding to
596
Annals of the
Missouri Botanical Garden
equal or greater than 50 nucleotides are ~ into
an appropriate data bank, e.g., GenBank/EMBL. The
accession numbers are provided before ons alion.
ong sequences (exceeding two pages) will not be rou-
tinely published.
Author accepts responsibility for establishing the ac-
curacy of information provided.
O
12. Keys
O
s are — and have been checked carefully for
consiste ney with the descriptions. Leads of each cou-
IL
plet are S
Dichotomous keys are indented
OO
Infraspecific taxa are keyed separately, not in species
keys
13. Literature Cited
The Literature Cited contains all references cited in
the text
All entries in the Literature Cited are cited in the text.
Spelling of author(s) name(s) and years of publication
have been double-checked.
O BB 0
All entries have been verified against — sources,
especially journal titles diacritical marks,
and spelling in languages pis than nadie sh.
Periodicals are listed authors last name,
d s). Year. ps title of article. Journal abbreviated
as in B-P-H/S. Volume: pages. No parenthetical part
tube ‘rs after lis numbers are given unless each
part is paginated separate
For more than one кз, his style is followed: au-
thors last name, initial(s), second author's initial(s),
last name & third author's initial(s), last
Books appear as follows: author's last name, initial(s).
Year. Full e Title (edited ii Editor), 3rd
ed., Vol. 2. Publisher, City of Publicatio
Citations i work “in prep.,” unpublished theses and
dissertations, and similar references
. accents,
d
as follows:
O
name.
O
O-
to inaccessible
sources have been eliminated or kept to a minimum.
They are not necessarily included in the Literature Cit-
ad
Illustrations
[]
Author(s) name(s) and figure number(s) are written in
pencil on the back of each figure or plate, on both
— and review copies
Seale bars appear on illustrations, photographs, and
maps
Magnifications/reductions are not indicated in
ions,
cap-
—
High- рЫ гога of the figures аге included
with each с A the manuscript.
Шо аге — to fit either column width (2
5/8 їп. or ca. 68 mm) or full page width (5% in. or са.
140 mm), or illustrations no larger than 11 X 16 in.
ERE
Volume 89, Number 4, pp. 453-596 of the ANNALS OF
was published on December 9, 2(
= 29 X 40 ст)
tions is 54 X
—
. [Maximum size for printed illustra-
X 8% in.] Oversized or delicate figures are
submitted as photostats.
"igures are numbered in Arabic numerals in the order
of their citation in the text. Parts of figures are labeled
with capital letters
[| Photographs are crisp black-and-white prints.
[] Figures are grouped into composite plates when pos-
sible: edges of р are abutted.
No stripping is inserted between plate or figure seg-
ments. [Printer will insert ‘ping
C] Edges of figures are squa
LJ Maps include reference to — and longitude and
are bounded by a fine border.
=
О Se anning electron micrographs are free of conspicuous
charging.
Axes on graphs are all labeled.
[] Captions provide all explanatory text. Captions are
separate from other text, one paragraph for each group
of figures, and following the style in current issues of
the Annals,
C] Symbols on maps are legible and reduction has been
taken into consideration.
ELECTRONIC ARTWORK—PRINTER’S SPECIFICATIONS
. We accept line art, halftones and color figures on disk
or via electronic delivery. Please include a printout or
email of the file directory that includes the file name(s),
size and kind of file.
2. You should discuss figure sizing specifications with the
journal prior to disk submission, and the figures on
disk should be the final size desired.
3. — пе files should be submitted in TIFF, EPS, or
PDF file format. Applic — that are acceptable are
Adobe Photoshop, Corel Draw, Adobe Illustrator and
Adobe Freehand, but please, if possible, “Save As’
uf, eps, or pdf file.
4. All line art should be scanned at 1200 d.p.i. All вту-
scale and color images should be scanned at 350 d.p
Color images should be submitted in CMYK mode, not
RGB or Indexed Color mode.
Include ra screen and printer fonts used within the
graphic file. Use only Adobe Type 1 postscript fonts.
True Type fonts and System “bitmap” fonts should not
be used.
If the files have been — резе indicate what
compression program has been used.
The disk submission should be ace a by a hard
copy of the figure that exactly matches what is on the
disk in the event that the disk is unusable.
ACCEPTABLE. ELECTRONIC
3% floppy disks; Syquest cartridges; EZ Flyer; ZIP
disks; JAZ disks; Magneto Optical disks; CD ROM. Art-
work may also be submitted via email attachment or FTP.
Please contact Karen Ridgway, e-mail address: kridgway@
allenpress.com, for more information.
ARTWORK MEDIA
THE MISSOURI BOTANICAL GARDEN
)02
ANNALS OF THE
MISSOURI BOTANICAL GARDEN
VOLUME 89
2002
Colophon
This volume of the ANNALS of the Missouri Botanical Garden has been set in APS Bodoni. The text
is set in 9 point type while the figure legends and literature cited sections are set in 8 point type.
This volume has been printed on 70# Vintage Gloss. This is an acid-free paper designed to have a
shelf-life of over 100 years. Vintage Gloss is manufactured by the Potlatch Paper Company.
Photographs used in the ANNALS are reproduced using 300 line screen halftones, The binding used
in the production of the ANNALS is a proprietary method known as Permanent Binding.
The ANNALS is printed and distributed by Allen Press, Inc. of Lawrence, Kansas 66044, U.S.A.
© Missouri Botanical Garden 2002
ISSN 0026-6493
VOLUME 89
2002
AGUILAR ENRIQUEZ, MARÍA DE LOURDES (see Silvia Romero Rangel.
Ezequiel Carlos Rojas Zenteno & Maria de Lourdes Aguilar Enríquez) — — 551
ALBACH, DIRK C. (see Steven J. Wagstaff, Michael J. Bayly, Philip J. Gar-
nock-Jones & Dirk C. Albach) |... 38
ALISCIONI, SANDRA S. Contribución a la Filogenia del Género Paspalum
(Poaceae: Panicoideae: Paniceae) 504
AL-SHEHBAZ, IHSAN A. (see Marcus Koch & Ihsan A. Al-Shehbaz) |... 88
BAYLY, MICHAEL J. (see Steven J. Wagstaff, Michael J. Bayly, Philip J. Gar-
nock-Jones € Dirk C. Albach) —. | 36
BRADFORD, JASON С. Molecular Phylogenetics and АВА ‘al Evolution
in Cunonieae (Cunoniaceae) : — 49]
CIALDELLA, ANA MARÍA & LILIANA MONICA GIUSSANI. Phylogenetic
Relationships of the Genus Piptochaetium (Poaceae, Pooideae, Stipeae):
Evidence from Morphological Data TM 305
CLAYTON, JASON R. (see Alan W. Meerow, Charles L. — Qin-Bao Li &
Jason R. Clayton) 000000 — 0
CRACRAFT, JOEL. The Seven Great Questions of Systematic Biology: An
Essential Foundation for Conservation and the Sustainable Use of Bio-
diversity 127
CROSBY, MARSHALL R. Statistical Summary of Some of the Activities in the
Missouri Botanical Garden Herbarium, 2001 303
DENHAM, SILVIA S., FERNANDO O. ZULOAGA & OSVALDO MORRONE.
Systematic Revision and Phylogeny of Paspalum Subgenus Ceresia (Po-
aceae: Panicoideae: Paniceae) — — < — < < =< — 337
DINERSTEIN, ERIC (see David M. Olson & Eric Dinerstein) 199
ERNEBERG, MARIANNE (see Richard N. Mack € Marianne Erneberg) |... 176
EZCURRA. CECILIA. El Género Justicia (Acanthaceae) en Sudamérica Aus-
JJJ. 225
FIALKOWSKI, CAROL (see Debra K. Moskovits, Carol Fialkowski, Gregory
M. Mueller & Timothy A. Sullivan) 153
FREIRE, SUSANA E., LILIANA KATINAS & GISELA SANCHO. Gochnatia
(Asteraceae, Mutisieae) and the Gochnatia Complex: Taxonomic bia
cations from Morphology — 524
GARNOCK-JONES, PHILIP J. (see Steven J. Wagstaff, Michael J. Bayly, Philip
J. Garnock-Jones & Dirk C. Albach) oa 38
GIUSSANI, LILIANA MÓNICA (see Ana María Cialdella & Liliana Mónica
Giussani) — ا 305
GOLDBLATT, PETER & JOHN C. MANNING. Evidence for Moth and But-
terfly Pollination in Gladiolus (Iridaceae—Crocoideae) — 110
GOLDBLATT, PETER & JOHN С. MANNING. Plant Diversity of the Cape
Region of Southern Africa — 281
GUY, CHARLES L. (see Alan W. Meerow, Charles L. Guy, Qin-Bao Li & Jason
R. Clayton)
HAUK, WARREN D. A Review of the Genera ee nia and Potamoganos
(Bignoniaceae) —
JOE, TONY (see Gary Paul Nabhan, Patrick "an & Tony Joe) |...
KATINAS, LILIANA (see Susana E. Freire, Liliana Katinas & Gisela Sanc ho)
KOCH, MARCUS & IHSAN A. AL-SHEHBAZ. Molecular Data Indicate Com-
plex Intra- and Intercontinental Differentiation of American Draba (Bras-
sicaceae) —
LI, QIN-BAO (see Alan W. Meerow, Charles L. Guy, Qin-Bao Li & Jason R.
Clayton) —
LOWRY, PORTER P, II (see Jun Wen, Porter P. Lowry ll, кыш. L. Walck &
Ki-Oug Yoo) — a
MACK, RICHARD N. € MARIANNE ERNEBERG. The United States Natu-
ralized Flora: Largely the Product of Deliberate Introductions ...
MANNING, JOHN C. (see Peter Goldblatt & John C. Manning) 0
MANNING, JOHN С. (see Peter Goldblatt & John С. Manning)
MEEROW, ALAN W., CHARLES L. GUY, QIN-BAO LI «€ JASON R. CLAY-
TON. Phylogeny of the Tribe Hymenocallideae NM eae) Based
on Morphology and Molecular Characters ........
MORRONE, OSVALDO (see Silvia S. Denham, Fernando O. Zuloaga & & Os-
valdo Morrone)
MOSKOVITS, DEBRA K., CAROL FIALKOWSKI, GREGORY M. MUELLER
& TIMOTHY A. SULLIVAN. Chicago Wilderness: a New Force in Urban
Conservation
MUELLER, GREGORY M. (see Debra K. Moskovits, Carol Fialkowski, Greg-
ory M. Mueller & Timothy A. Sullivan)
NABHAN, GARY PAUL, PATRICK PYNES & TONY JOE. Safeguarding Spe-
cies, Languages, and Cultures in the Time of Diversity Loss: From the
Colorado Plateau to Global Hotspots —
OLSON, DAVID M. & ERIC DINERSTEIN. The Global 200: Priority Ecore-
gions for Global Conservation |...
PIMM, STUART L. The Dodo Went Extinct (and Other Ecological Myths) .. —
PORTER, J. MARK (see Victor W. Steinmann & J. Mark Porter) .......
PYNES, PATRICK (see Gary Paul Nabhan, Patrick Pynes & Tony Joe)
RAZAFIMANDIMBISON, SYLVAIN G. A Systematic Revision of Breonia (Ru-
ylaceae—Naucleeae)
RICHARDSON, P. MICK (see George E. Schatz & P. Mick Ric — —
ROJAS ZENTENO, EZEQUIEL CARLOS (see Silvia Romero Rangel, Eze аи
Carlos Rojas Zenteno & María de Lourdes Aguilar Enriquez) —
524
88
400
414
176
110
281
400
ROMERO RANGEL, SILVIA, EZEQUIEL CARLOS ROJAS ZENTENO &
MARIA DE LOURDES AGUILAR ENRIQUEZ. El Género Quercus (Fa-
gaceae) en el Estado de México
SANCHO, GISELA (see Susana E. Freire, Liliana Katinas & Gisela Sancho)
SCHATZ, GEORGE К. Taxonomy and Herbaria in Service of Plant Conser-
vation: Lessons from Madagascar’s Endemic Families
SCHATZ, GEORGE К. & Р. MICK RICHARDSON. Conservation, the 47th
Annual Systematics Symposium of the Missouri Botanical Garden. Intro-
duction: Conservation and the Future of Life
SCHNEIDER, JULIO V., ULF SWENSON & GEORG ZIZKA. Phylogenetic
Reconstruction of the Neotropical Family Quiinaceae (Malpighiales)
Based on Morphology with Remarks on the Evolution of an Androdioe-
cious Sex Distribution
STEINMANN, VICTOR W. & J. MARK PORTER. Phylogenetic Relationships
in Euphorbieae (Euphorbiaceae) Based on ITS and ndhF Sequence Data
SULLIVAN, TIMOTHY A. (see Debra K. Moskovits, Carol Fialkowski, Gregory
M. Mueller & Timothy A. Sullivan)
SWENSON, ULF (see Julio V. Schneider, Ulf Swenson & Georg Zizka) ~~
VAN DER WERFF, HENK. A Synopsis of Ocotea (Lauraceae) in Central Amer-
ica and Southern Mexico
WAGSTAFF, STEVEN J., MICHAEL J. BAYLY, PHILIP J. GARNOCK-JONES
& DIRK С. ALBACH. Classification, Origin, and Diversification of the
New Zealand Hebes (Scrophulariaceae)
WALCK, JEFFREY L. (see Jun Wen, Porter P. Lowry II, Jeffrey L. Walck &
Ki-Oug Yoo)
WEN. JUN, PORTER P. LOWRY II, JEFFREY L. WALCK & KI-OUG YOO.
Phylogenetic and Biogeographic Diversification in Osmorhiza (Apiaceae)
YOO, KI-OUG (see Jun Wen, Porter P. Lowry П, Jeffrey L. Walck € Ki-Oug
00
ZIZKA, GEORG (see Julio V. Schneider, Ulf Swenson & Georg Zizka) —
ZULOAGA, FERNANDO О. (see Silvia S. Denham, Fernando O. Zuloaga &
Osvaldo Morrone)
їл
yu
—
524
145
125
64
(EE
L m]
— |
ANNALS OF THE MISSOURI BOTANICAL GARDEN
|
AND MISSOURI BOTANICAL GARDEN ANNUAL REPORT |
ARE NOW AVAILABLE IN JSTOR! |
|
JSTOR’, a not-for-profit organization, is an important endeavor
dedicated to helping the scholarly community take advantage of advances
in electronic technologies. The JSTOR Archive makes available the
complete backfiles of Annals of The Missouri Botanical Garden,
Missouri Botanical Garden Annual Report (1890-1912), and more than 100
other scholarly journals to researchers through participating libraries.
JSTOR has recently announced the release of its Ecology & Botany
Collection. Developed in cooperation with the Ecological Society of America,
the Collection includes the archives of twenty-nine journals, offering
online access to more than 1 million pages of important historic literature.
The Ecology & Botany Collection
American Journal of Botany International Journal of Plant Sciences
American Midland Naturalist Journal of Animal Ecology
American Naturalist Journal of Applied Ecology
Annals of the Missouri Botanical Garden Journal of Biogeography |
Annual Review of Ecology and Systematics Journal of Ecology
Biotropica Journal of Tropical Ecology
Brittonia Journal of The Torrey Botanical Society
Conservation Biology | Limnology and Oceanography
Diversity and Distributions Missouri Botanical Gaiden Annual Ene |
Ecological Applications New Phytologist
Ecological Monographs . Paleobiology
Ecology Quarterly Review of poer
Evolution Systematic Biology
Functional Ecology Systematic Botany
Global Ecology and Biogeography Letters
Information regarding JSTOR is available at http://www.jstor.org
120 Fifth Avenue, New York, NY 10011 —
CONTENTS
Phylogenetic Relationships in Euphorbieae (Euphorbiaceae) Based on ITS and ndhF
Sequence Data Victor W. Steinmann & J. Mark Porter
Molecular Phylogenetics and Morphological Evolution in Cunonieae (Cunoniaceae)
C. Bradford
Contribución a la Filogenia del Género Paspalum (Poaceae: Panicoideae: Paniceae)
Sandra S. Aliscioni
Gochnatia (Asteraceae, Mutisieae) and the Gochnatia Complex: Taxonomic Impli-
cations from Morphology ...... Susana E. Freire, Liliana Katinas & Gisela Sancho
El Género Quercus (Fagaceae) en el Estado de México ....... - Silvia Romero Rangel,
— Ezequiel Carlos Rojas Zenteno y María de Lourdes Aguilar Enríquez
Checklist T Rites
Index to Volume 89
_ Cover illustration. — Breonia richardsonii Razafim.. d Es ik Misi
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11